1 /**************************************************************************
3 Copyright (c) 2007-2009, Chelsio Inc.
6 Redistribution and use in source and binary forms, with or without
7 modification, are permitted provided that the following conditions are met:
9 1. Redistributions of source code must retain the above copyright notice,
10 this list of conditions and the following disclaimer.
12 2. Neither the name of the Chelsio Corporation nor the names of its
13 contributors may be used to endorse or promote products derived from
14 this software without specific prior written permission.
16 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
17 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19 ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
20 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
21 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
22 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
23 INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
24 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
25 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
26 POSSIBILITY OF SUCH DAMAGE.
28 ***************************************************************************/
30 #include <sys/cdefs.h>
31 __FBSDID("$FreeBSD$");
33 #include "opt_inet6.h"
36 #include <sys/param.h>
37 #include <sys/systm.h>
38 #include <sys/kernel.h>
39 #include <sys/module.h>
42 #include <machine/bus.h>
43 #include <machine/resource.h>
44 #include <sys/bus_dma.h>
46 #include <sys/queue.h>
47 #include <sys/sysctl.h>
48 #include <sys/taskqueue.h>
52 #include <sys/sched.h>
54 #include <sys/systm.h>
55 #include <sys/syslog.h>
56 #include <sys/socket.h>
57 #include <sys/sglist.h>
60 #include <net/ethernet.h>
62 #include <net/if_vlan_var.h>
64 #include <netinet/in_systm.h>
65 #include <netinet/in.h>
66 #include <netinet/ip.h>
67 #include <netinet/ip6.h>
68 #include <netinet/tcp.h>
70 #include <dev/pci/pcireg.h>
71 #include <dev/pci/pcivar.h>
76 #include <cxgb_include.h>
80 int multiq_tx_enable = 1;
83 CTASSERT(NUM_CPL_HANDLERS >= NUM_CPL_CMDS);
86 extern struct sysctl_oid_list sysctl__hw_cxgb_children;
87 int cxgb_txq_buf_ring_size = TX_ETH_Q_SIZE;
88 TUNABLE_INT("hw.cxgb.txq_mr_size", &cxgb_txq_buf_ring_size);
89 SYSCTL_INT(_hw_cxgb, OID_AUTO, txq_mr_size, CTLFLAG_RDTUN, &cxgb_txq_buf_ring_size, 0,
90 "size of per-queue mbuf ring");
92 static int cxgb_tx_coalesce_force = 0;
93 TUNABLE_INT("hw.cxgb.tx_coalesce_force", &cxgb_tx_coalesce_force);
94 SYSCTL_INT(_hw_cxgb, OID_AUTO, tx_coalesce_force, CTLFLAG_RW,
95 &cxgb_tx_coalesce_force, 0,
96 "coalesce small packets into a single work request regardless of ring state");
98 #define COALESCE_START_DEFAULT TX_ETH_Q_SIZE>>1
99 #define COALESCE_START_MAX (TX_ETH_Q_SIZE-(TX_ETH_Q_SIZE>>3))
100 #define COALESCE_STOP_DEFAULT TX_ETH_Q_SIZE>>2
101 #define COALESCE_STOP_MIN TX_ETH_Q_SIZE>>5
102 #define TX_RECLAIM_DEFAULT TX_ETH_Q_SIZE>>5
103 #define TX_RECLAIM_MAX TX_ETH_Q_SIZE>>2
104 #define TX_RECLAIM_MIN TX_ETH_Q_SIZE>>6
107 static int cxgb_tx_coalesce_enable_start = COALESCE_START_DEFAULT;
108 TUNABLE_INT("hw.cxgb.tx_coalesce_enable_start",
109 &cxgb_tx_coalesce_enable_start);
110 SYSCTL_INT(_hw_cxgb, OID_AUTO, tx_coalesce_enable_start, CTLFLAG_RW,
111 &cxgb_tx_coalesce_enable_start, 0,
112 "coalesce enable threshold");
113 static int cxgb_tx_coalesce_enable_stop = COALESCE_STOP_DEFAULT;
114 TUNABLE_INT("hw.cxgb.tx_coalesce_enable_stop", &cxgb_tx_coalesce_enable_stop);
115 SYSCTL_INT(_hw_cxgb, OID_AUTO, tx_coalesce_enable_stop, CTLFLAG_RW,
116 &cxgb_tx_coalesce_enable_stop, 0,
117 "coalesce disable threshold");
118 static int cxgb_tx_reclaim_threshold = TX_RECLAIM_DEFAULT;
119 TUNABLE_INT("hw.cxgb.tx_reclaim_threshold", &cxgb_tx_reclaim_threshold);
120 SYSCTL_INT(_hw_cxgb, OID_AUTO, tx_reclaim_threshold, CTLFLAG_RW,
121 &cxgb_tx_reclaim_threshold, 0,
122 "tx cleaning minimum threshold");
125 * XXX don't re-enable this until TOE stops assuming
128 static int recycle_enable = 0;
130 extern int cxgb_use_16k_clusters;
131 extern int nmbjumbop;
132 extern int nmbjumbo9;
133 extern int nmbjumbo16;
137 #define SGE_RX_SM_BUF_SIZE 1536
138 #define SGE_RX_DROP_THRES 16
139 #define SGE_RX_COPY_THRES 128
142 * Period of the Tx buffer reclaim timer. This timer does not need to run
143 * frequently as Tx buffers are usually reclaimed by new Tx packets.
145 #define TX_RECLAIM_PERIOD (hz >> 1)
148 * Values for sge_txq.flags
151 TXQ_RUNNING = 1 << 0, /* fetch engine is running */
152 TXQ_LAST_PKT_DB = 1 << 1, /* last packet rang the doorbell */
156 uint64_t flit[TX_DESC_FLITS];
166 struct rsp_desc { /* response queue descriptor */
167 struct rss_header rss_hdr;
170 uint8_t imm_data[47];
174 #define RX_SW_DESC_MAP_CREATED (1 << 0)
175 #define TX_SW_DESC_MAP_CREATED (1 << 1)
176 #define RX_SW_DESC_INUSE (1 << 3)
177 #define TX_SW_DESC_MAPPED (1 << 4)
179 #define RSPQ_NSOP_NEOP G_RSPD_SOP_EOP(0)
180 #define RSPQ_EOP G_RSPD_SOP_EOP(F_RSPD_EOP)
181 #define RSPQ_SOP G_RSPD_SOP_EOP(F_RSPD_SOP)
182 #define RSPQ_SOP_EOP G_RSPD_SOP_EOP(F_RSPD_SOP|F_RSPD_EOP)
184 struct tx_sw_desc { /* SW state per Tx descriptor */
190 struct rx_sw_desc { /* SW state per Rx descriptor */
203 struct refill_fl_cb_arg {
205 bus_dma_segment_t seg;
211 * Maps a number of flits to the number of Tx descriptors that can hold them.
214 * desc = 1 + (flits - 2) / (WR_FLITS - 1).
216 * HW allows up to 4 descriptors to be combined into a WR.
218 static uint8_t flit_desc_map[] = {
220 #if SGE_NUM_GENBITS == 1
221 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
222 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
223 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
224 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4
225 #elif SGE_NUM_GENBITS == 2
226 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
227 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
228 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
229 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4,
231 # error "SGE_NUM_GENBITS must be 1 or 2"
235 #define TXQ_LOCK_ASSERT(qs) mtx_assert(&(qs)->lock, MA_OWNED)
236 #define TXQ_TRYLOCK(qs) mtx_trylock(&(qs)->lock)
237 #define TXQ_LOCK(qs) mtx_lock(&(qs)->lock)
238 #define TXQ_UNLOCK(qs) mtx_unlock(&(qs)->lock)
239 #define TXQ_RING_EMPTY(qs) drbr_empty((qs)->port->ifp, (qs)->txq[TXQ_ETH].txq_mr)
240 #define TXQ_RING_NEEDS_ENQUEUE(qs) \
241 drbr_needs_enqueue((qs)->port->ifp, (qs)->txq[TXQ_ETH].txq_mr)
242 #define TXQ_RING_FLUSH(qs) drbr_flush((qs)->port->ifp, (qs)->txq[TXQ_ETH].txq_mr)
243 #define TXQ_RING_DEQUEUE_COND(qs, func, arg) \
244 drbr_dequeue_cond((qs)->port->ifp, (qs)->txq[TXQ_ETH].txq_mr, func, arg)
245 #define TXQ_RING_DEQUEUE(qs) \
246 drbr_dequeue((qs)->port->ifp, (qs)->txq[TXQ_ETH].txq_mr)
250 static void sge_timer_cb(void *arg);
251 static void sge_timer_reclaim(void *arg, int ncount);
252 static void sge_txq_reclaim_handler(void *arg, int ncount);
253 static void cxgb_start_locked(struct sge_qset *qs);
256 * XXX need to cope with bursty scheduling by looking at a wider
257 * window than we are now for determining the need for coalescing
260 static __inline uint64_t
261 check_pkt_coalesce(struct sge_qset *qs)
267 if (__predict_false(cxgb_tx_coalesce_force))
269 txq = &qs->txq[TXQ_ETH];
270 sc = qs->port->adapter;
271 fill = &sc->tunq_fill[qs->idx];
273 if (cxgb_tx_coalesce_enable_start > COALESCE_START_MAX)
274 cxgb_tx_coalesce_enable_start = COALESCE_START_MAX;
275 if (cxgb_tx_coalesce_enable_stop < COALESCE_STOP_MIN)
276 cxgb_tx_coalesce_enable_start = COALESCE_STOP_MIN;
278 * if the hardware transmit queue is more than 1/8 full
279 * we mark it as coalescing - we drop back from coalescing
280 * when we go below 1/32 full and there are no packets enqueued,
281 * this provides us with some degree of hysteresis
283 if (*fill != 0 && (txq->in_use <= cxgb_tx_coalesce_enable_stop) &&
284 TXQ_RING_EMPTY(qs) && (qs->coalescing == 0))
286 else if (*fill == 0 && (txq->in_use >= cxgb_tx_coalesce_enable_start))
289 return (sc->tunq_coalesce);
294 set_wr_hdr(struct work_request_hdr *wrp, uint32_t wr_hi, uint32_t wr_lo)
297 #if _BYTE_ORDER == _LITTLE_ENDIAN
299 wr_hilo |= (((uint64_t)wr_lo)<<32);
302 wr_hilo |= (((uint64_t)wr_hi)<<32);
304 wrp->wrh_hilo = wr_hilo;
308 set_wr_hdr(struct work_request_hdr *wrp, uint32_t wr_hi, uint32_t wr_lo)
317 struct coalesce_info {
323 coalesce_check(struct mbuf *m, void *arg)
325 struct coalesce_info *ci = arg;
326 int *count = &ci->count;
327 int *nbytes = &ci->nbytes;
329 if ((*nbytes == 0) || ((*nbytes + m->m_len <= 10500) &&
330 (*count < 7) && (m->m_next == NULL))) {
339 cxgb_dequeue(struct sge_qset *qs)
341 struct mbuf *m, *m_head, *m_tail;
342 struct coalesce_info ci;
345 if (check_pkt_coalesce(qs) == 0)
346 return TXQ_RING_DEQUEUE(qs);
348 m_head = m_tail = NULL;
349 ci.count = ci.nbytes = 0;
351 m = TXQ_RING_DEQUEUE_COND(qs, coalesce_check, &ci);
352 if (m_head == NULL) {
354 } else if (m != NULL) {
355 m_tail->m_nextpkt = m;
360 panic("trying to coalesce %d packets in to one WR", ci.count);
365 * reclaim_completed_tx - reclaims completed Tx descriptors
366 * @adapter: the adapter
367 * @q: the Tx queue to reclaim completed descriptors from
369 * Reclaims Tx descriptors that the SGE has indicated it has processed,
370 * and frees the associated buffers if possible. Called with the Tx
374 reclaim_completed_tx(struct sge_qset *qs, int reclaim_min, int queue)
376 struct sge_txq *q = &qs->txq[queue];
377 int reclaim = desc_reclaimable(q);
379 if ((cxgb_tx_reclaim_threshold > TX_RECLAIM_MAX) ||
380 (cxgb_tx_reclaim_threshold < TX_RECLAIM_MIN))
381 cxgb_tx_reclaim_threshold = TX_RECLAIM_DEFAULT;
383 if (reclaim < reclaim_min)
386 mtx_assert(&qs->lock, MA_OWNED);
388 t3_free_tx_desc(qs, reclaim, queue);
389 q->cleaned += reclaim;
390 q->in_use -= reclaim;
392 if (isset(&qs->txq_stopped, TXQ_ETH))
393 clrbit(&qs->txq_stopped, TXQ_ETH);
399 * should_restart_tx - are there enough resources to restart a Tx queue?
402 * Checks if there are enough descriptors to restart a suspended Tx queue.
405 should_restart_tx(const struct sge_txq *q)
407 unsigned int r = q->processed - q->cleaned;
409 return q->in_use - r < (q->size >> 1);
413 * t3_sge_init - initialize SGE
415 * @p: the SGE parameters
417 * Performs SGE initialization needed every time after a chip reset.
418 * We do not initialize any of the queue sets here, instead the driver
419 * top-level must request those individually. We also do not enable DMA
420 * here, that should be done after the queues have been set up.
423 t3_sge_init(adapter_t *adap, struct sge_params *p)
427 ups = 0; /* = ffs(pci_resource_len(adap->pdev, 2) >> 12); */
429 ctrl = F_DROPPKT | V_PKTSHIFT(2) | F_FLMODE | F_AVOIDCQOVFL |
430 F_CQCRDTCTRL | F_CONGMODE | F_TNLFLMODE | F_FATLPERREN |
431 V_HOSTPAGESIZE(PAGE_SHIFT - 11) | F_BIGENDIANINGRESS |
432 V_USERSPACESIZE(ups ? ups - 1 : 0) | F_ISCSICOALESCING;
433 #if SGE_NUM_GENBITS == 1
434 ctrl |= F_EGRGENCTRL;
436 if (adap->params.rev > 0) {
437 if (!(adap->flags & (USING_MSIX | USING_MSI)))
438 ctrl |= F_ONEINTMULTQ | F_OPTONEINTMULTQ;
440 t3_write_reg(adap, A_SG_CONTROL, ctrl);
441 t3_write_reg(adap, A_SG_EGR_RCQ_DRB_THRSH, V_HIRCQDRBTHRSH(512) |
442 V_LORCQDRBTHRSH(512));
443 t3_write_reg(adap, A_SG_TIMER_TICK, core_ticks_per_usec(adap) / 10);
444 t3_write_reg(adap, A_SG_CMDQ_CREDIT_TH, V_THRESHOLD(32) |
445 V_TIMEOUT(200 * core_ticks_per_usec(adap)));
446 t3_write_reg(adap, A_SG_HI_DRB_HI_THRSH,
447 adap->params.rev < T3_REV_C ? 1000 : 500);
448 t3_write_reg(adap, A_SG_HI_DRB_LO_THRSH, 256);
449 t3_write_reg(adap, A_SG_LO_DRB_HI_THRSH, 1000);
450 t3_write_reg(adap, A_SG_LO_DRB_LO_THRSH, 256);
451 t3_write_reg(adap, A_SG_OCO_BASE, V_BASE1(0xfff));
452 t3_write_reg(adap, A_SG_DRB_PRI_THRESH, 63 * 1024);
457 * sgl_len - calculates the size of an SGL of the given capacity
458 * @n: the number of SGL entries
460 * Calculates the number of flits needed for a scatter/gather list that
461 * can hold the given number of entries.
463 static __inline unsigned int
464 sgl_len(unsigned int n)
466 return ((3 * n) / 2 + (n & 1));
470 * get_imm_packet - return the next ingress packet buffer from a response
471 * @resp: the response descriptor containing the packet data
473 * Return a packet containing the immediate data of the given response.
476 get_imm_packet(adapter_t *sc, const struct rsp_desc *resp, struct mbuf *m)
479 if (resp->rss_hdr.opcode == CPL_RX_DATA) {
480 const struct cpl_rx_data *cpl = (const void *)&resp->imm_data[0];
481 m->m_len = sizeof(*cpl) + ntohs(cpl->len);
482 } else if (resp->rss_hdr.opcode == CPL_RX_PKT) {
483 const struct cpl_rx_pkt *cpl = (const void *)&resp->imm_data[0];
484 m->m_len = sizeof(*cpl) + ntohs(cpl->len);
486 m->m_len = IMMED_PKT_SIZE;
487 m->m_ext.ext_buf = NULL;
488 m->m_ext.ext_type = 0;
489 memcpy(mtod(m, uint8_t *), resp->imm_data, m->m_len);
493 static __inline u_int
494 flits_to_desc(u_int n)
496 return (flit_desc_map[n]);
499 #define SGE_PARERR (F_CPPARITYERROR | F_OCPARITYERROR | F_RCPARITYERROR | \
500 F_IRPARITYERROR | V_ITPARITYERROR(M_ITPARITYERROR) | \
501 V_FLPARITYERROR(M_FLPARITYERROR) | F_LODRBPARITYERROR | \
502 F_HIDRBPARITYERROR | F_LORCQPARITYERROR | \
504 #define SGE_FRAMINGERR (F_UC_REQ_FRAMINGERROR | F_R_REQ_FRAMINGERROR)
505 #define SGE_FATALERR (SGE_PARERR | SGE_FRAMINGERR | F_RSPQCREDITOVERFOW | \
509 * t3_sge_err_intr_handler - SGE async event interrupt handler
510 * @adapter: the adapter
512 * Interrupt handler for SGE asynchronous (non-data) events.
515 t3_sge_err_intr_handler(adapter_t *adapter)
517 unsigned int v, status;
519 status = t3_read_reg(adapter, A_SG_INT_CAUSE);
520 if (status & SGE_PARERR)
521 CH_ALERT(adapter, "SGE parity error (0x%x)\n",
522 status & SGE_PARERR);
523 if (status & SGE_FRAMINGERR)
524 CH_ALERT(adapter, "SGE framing error (0x%x)\n",
525 status & SGE_FRAMINGERR);
526 if (status & F_RSPQCREDITOVERFOW)
527 CH_ALERT(adapter, "SGE response queue credit overflow\n");
529 if (status & F_RSPQDISABLED) {
530 v = t3_read_reg(adapter, A_SG_RSPQ_FL_STATUS);
533 "packet delivered to disabled response queue (0x%x)\n",
534 (v >> S_RSPQ0DISABLED) & 0xff);
537 t3_write_reg(adapter, A_SG_INT_CAUSE, status);
538 if (status & SGE_FATALERR)
539 t3_fatal_err(adapter);
543 t3_sge_prep(adapter_t *adap, struct sge_params *p)
545 int i, nqsets, fl_q_size, jumbo_q_size, use_16k, jumbo_buf_size;
547 nqsets = min(SGE_QSETS / adap->params.nports, mp_ncpus);
548 nqsets *= adap->params.nports;
550 fl_q_size = min(nmbclusters/(3*nqsets), FL_Q_SIZE);
552 while (!powerof2(fl_q_size))
555 use_16k = cxgb_use_16k_clusters != -1 ? cxgb_use_16k_clusters :
558 #if __FreeBSD_version >= 700111
560 jumbo_q_size = min(nmbjumbo16/(3*nqsets), JUMBO_Q_SIZE);
561 jumbo_buf_size = MJUM16BYTES;
563 jumbo_q_size = min(nmbjumbo9/(3*nqsets), JUMBO_Q_SIZE);
564 jumbo_buf_size = MJUM9BYTES;
567 jumbo_q_size = min(nmbjumbop/(3*nqsets), JUMBO_Q_SIZE);
568 jumbo_buf_size = MJUMPAGESIZE;
570 while (!powerof2(jumbo_q_size))
573 if (fl_q_size < (FL_Q_SIZE / 4) || jumbo_q_size < (JUMBO_Q_SIZE / 2))
574 device_printf(adap->dev,
575 "Insufficient clusters and/or jumbo buffers.\n");
577 p->max_pkt_size = jumbo_buf_size - sizeof(struct cpl_rx_data);
579 for (i = 0; i < SGE_QSETS; ++i) {
580 struct qset_params *q = p->qset + i;
582 if (adap->params.nports > 2) {
583 q->coalesce_usecs = 50;
586 q->coalesce_usecs = 10;
588 q->coalesce_usecs = 5;
592 q->rspq_size = RSPQ_Q_SIZE;
593 q->fl_size = fl_q_size;
594 q->jumbo_size = jumbo_q_size;
595 q->jumbo_buf_size = jumbo_buf_size;
596 q->txq_size[TXQ_ETH] = TX_ETH_Q_SIZE;
597 q->txq_size[TXQ_OFLD] = is_offload(adap) ? TX_OFLD_Q_SIZE : 16;
598 q->txq_size[TXQ_CTRL] = TX_CTRL_Q_SIZE;
604 t3_sge_alloc(adapter_t *sc)
607 /* The parent tag. */
608 if (bus_dma_tag_create( bus_get_dma_tag(sc->dev),/* PCI parent */
609 1, 0, /* algnmnt, boundary */
610 BUS_SPACE_MAXADDR, /* lowaddr */
611 BUS_SPACE_MAXADDR, /* highaddr */
612 NULL, NULL, /* filter, filterarg */
613 BUS_SPACE_MAXSIZE_32BIT,/* maxsize */
614 BUS_SPACE_UNRESTRICTED, /* nsegments */
615 BUS_SPACE_MAXSIZE_32BIT,/* maxsegsize */
617 NULL, NULL, /* lock, lockarg */
619 device_printf(sc->dev, "Cannot allocate parent DMA tag\n");
624 * DMA tag for normal sized RX frames
626 if (bus_dma_tag_create(sc->parent_dmat, MCLBYTES, 0, BUS_SPACE_MAXADDR,
627 BUS_SPACE_MAXADDR, NULL, NULL, MCLBYTES, 1,
628 MCLBYTES, BUS_DMA_ALLOCNOW, NULL, NULL, &sc->rx_dmat)) {
629 device_printf(sc->dev, "Cannot allocate RX DMA tag\n");
634 * DMA tag for jumbo sized RX frames.
636 if (bus_dma_tag_create(sc->parent_dmat, MJUM16BYTES, 0, BUS_SPACE_MAXADDR,
637 BUS_SPACE_MAXADDR, NULL, NULL, MJUM16BYTES, 1, MJUM16BYTES,
638 BUS_DMA_ALLOCNOW, NULL, NULL, &sc->rx_jumbo_dmat)) {
639 device_printf(sc->dev, "Cannot allocate RX jumbo DMA tag\n");
644 * DMA tag for TX frames.
646 if (bus_dma_tag_create(sc->parent_dmat, 1, 0, BUS_SPACE_MAXADDR,
647 BUS_SPACE_MAXADDR, NULL, NULL, TX_MAX_SIZE, TX_MAX_SEGS,
648 TX_MAX_SIZE, BUS_DMA_ALLOCNOW,
649 NULL, NULL, &sc->tx_dmat)) {
650 device_printf(sc->dev, "Cannot allocate TX DMA tag\n");
658 t3_sge_free(struct adapter * sc)
661 if (sc->tx_dmat != NULL)
662 bus_dma_tag_destroy(sc->tx_dmat);
664 if (sc->rx_jumbo_dmat != NULL)
665 bus_dma_tag_destroy(sc->rx_jumbo_dmat);
667 if (sc->rx_dmat != NULL)
668 bus_dma_tag_destroy(sc->rx_dmat);
670 if (sc->parent_dmat != NULL)
671 bus_dma_tag_destroy(sc->parent_dmat);
677 t3_update_qset_coalesce(struct sge_qset *qs, const struct qset_params *p)
680 qs->rspq.holdoff_tmr = max(p->coalesce_usecs * 10, 1U);
681 qs->rspq.polling = 0 /* p->polling */;
684 #if !defined(__i386__) && !defined(__amd64__)
686 refill_fl_cb(void *arg, bus_dma_segment_t *segs, int nseg, int error)
688 struct refill_fl_cb_arg *cb_arg = arg;
690 cb_arg->error = error;
691 cb_arg->seg = segs[0];
697 * refill_fl - refill an SGE free-buffer list
698 * @sc: the controller softc
699 * @q: the free-list to refill
700 * @n: the number of new buffers to allocate
702 * (Re)populate an SGE free-buffer list with up to @n new packet buffers.
703 * The caller must assure that @n does not exceed the queue's capacity.
706 refill_fl(adapter_t *sc, struct sge_fl *q, int n)
708 struct rx_sw_desc *sd = &q->sdesc[q->pidx];
709 struct rx_desc *d = &q->desc[q->pidx];
710 struct refill_fl_cb_arg cb_arg;
718 * We allocate an uninitialized mbuf + cluster, mbuf is
719 * initialized after rx.
721 if (q->zone == zone_pack) {
722 if ((m = m_getcl(M_NOWAIT, MT_NOINIT, M_PKTHDR)) == NULL)
724 cl = m->m_ext.ext_buf;
726 if ((cl = m_cljget(NULL, M_NOWAIT, q->buf_size)) == NULL)
728 if ((m = m_gethdr(M_NOWAIT, MT_NOINIT)) == NULL) {
729 uma_zfree(q->zone, cl);
733 if ((sd->flags & RX_SW_DESC_MAP_CREATED) == 0) {
734 if ((err = bus_dmamap_create(q->entry_tag, 0, &sd->map))) {
735 log(LOG_WARNING, "bus_dmamap_create failed %d\n", err);
736 uma_zfree(q->zone, cl);
739 sd->flags |= RX_SW_DESC_MAP_CREATED;
741 #if !defined(__i386__) && !defined(__amd64__)
742 err = bus_dmamap_load(q->entry_tag, sd->map,
743 cl, q->buf_size, refill_fl_cb, &cb_arg, 0);
745 if (err != 0 || cb_arg.error) {
746 if (q->zone == zone_pack)
747 uma_zfree(q->zone, cl);
752 cb_arg.seg.ds_addr = pmap_kextract((vm_offset_t)cl);
754 sd->flags |= RX_SW_DESC_INUSE;
757 d->addr_lo = htobe32(cb_arg.seg.ds_addr & 0xffffffff);
758 d->addr_hi = htobe32(((uint64_t)cb_arg.seg.ds_addr >>32) & 0xffffffff);
759 d->len_gen = htobe32(V_FLD_GEN1(q->gen));
760 d->gen2 = htobe32(V_FLD_GEN2(q->gen));
765 if (++q->pidx == q->size) {
776 if (q->db_pending >= 32) {
778 t3_write_reg(sc, A_SG_KDOORBELL, V_EGRCNTX(q->cntxt_id));
784 * free_rx_bufs - free the Rx buffers on an SGE free list
785 * @sc: the controle softc
786 * @q: the SGE free list to clean up
788 * Release the buffers on an SGE free-buffer Rx queue. HW fetching from
789 * this queue should be stopped before calling this function.
792 free_rx_bufs(adapter_t *sc, struct sge_fl *q)
794 u_int cidx = q->cidx;
796 while (q->credits--) {
797 struct rx_sw_desc *d = &q->sdesc[cidx];
799 if (d->flags & RX_SW_DESC_INUSE) {
800 bus_dmamap_unload(q->entry_tag, d->map);
801 bus_dmamap_destroy(q->entry_tag, d->map);
802 if (q->zone == zone_pack) {
803 m_init(d->m, zone_pack, MCLBYTES,
804 M_NOWAIT, MT_DATA, M_EXT);
805 uma_zfree(zone_pack, d->m);
807 m_init(d->m, zone_mbuf, MLEN,
808 M_NOWAIT, MT_DATA, 0);
809 uma_zfree(zone_mbuf, d->m);
810 uma_zfree(q->zone, d->rxsd_cl);
816 if (++cidx == q->size)
822 __refill_fl(adapter_t *adap, struct sge_fl *fl)
824 refill_fl(adap, fl, min(16U, fl->size - fl->credits));
828 __refill_fl_lt(adapter_t *adap, struct sge_fl *fl, int max)
830 uint32_t reclaimable = fl->size - fl->credits;
833 refill_fl(adap, fl, min(max, reclaimable));
837 * recycle_rx_buf - recycle a receive buffer
838 * @adapter: the adapter
839 * @q: the SGE free list
840 * @idx: index of buffer to recycle
842 * Recycles the specified buffer on the given free list by adding it at
843 * the next available slot on the list.
846 recycle_rx_buf(adapter_t *adap, struct sge_fl *q, unsigned int idx)
848 struct rx_desc *from = &q->desc[idx];
849 struct rx_desc *to = &q->desc[q->pidx];
851 q->sdesc[q->pidx] = q->sdesc[idx];
852 to->addr_lo = from->addr_lo; // already big endian
853 to->addr_hi = from->addr_hi; // likewise
854 wmb(); /* necessary ? */
855 to->len_gen = htobe32(V_FLD_GEN1(q->gen));
856 to->gen2 = htobe32(V_FLD_GEN2(q->gen));
859 if (++q->pidx == q->size) {
863 t3_write_reg(adap, A_SG_KDOORBELL, V_EGRCNTX(q->cntxt_id));
867 alloc_ring_cb(void *arg, bus_dma_segment_t *segs, int nsegs, int error)
872 *addr = segs[0].ds_addr;
876 alloc_ring(adapter_t *sc, size_t nelem, size_t elem_size, size_t sw_size,
877 bus_addr_t *phys, void *desc, void *sdesc, bus_dma_tag_t *tag,
878 bus_dmamap_t *map, bus_dma_tag_t parent_entry_tag, bus_dma_tag_t *entry_tag)
880 size_t len = nelem * elem_size;
885 if ((err = bus_dma_tag_create(sc->parent_dmat, PAGE_SIZE, 0,
886 BUS_SPACE_MAXADDR_32BIT,
887 BUS_SPACE_MAXADDR, NULL, NULL, len, 1,
888 len, 0, NULL, NULL, tag)) != 0) {
889 device_printf(sc->dev, "Cannot allocate descriptor tag\n");
893 if ((err = bus_dmamem_alloc(*tag, (void **)&p, BUS_DMA_NOWAIT,
895 device_printf(sc->dev, "Cannot allocate descriptor memory\n");
899 bus_dmamap_load(*tag, *map, p, len, alloc_ring_cb, phys, 0);
904 len = nelem * sw_size;
905 s = malloc(len, M_DEVBUF, M_WAITOK|M_ZERO);
908 if (parent_entry_tag == NULL)
911 if ((err = bus_dma_tag_create(parent_entry_tag, 1, 0,
912 BUS_SPACE_MAXADDR, BUS_SPACE_MAXADDR,
913 NULL, NULL, TX_MAX_SIZE, TX_MAX_SEGS,
914 TX_MAX_SIZE, BUS_DMA_ALLOCNOW,
915 NULL, NULL, entry_tag)) != 0) {
916 device_printf(sc->dev, "Cannot allocate descriptor entry tag\n");
923 sge_slow_intr_handler(void *arg, int ncount)
927 t3_slow_intr_handler(sc);
928 t3_write_reg(sc, A_PL_INT_ENABLE0, sc->slow_intr_mask);
929 (void) t3_read_reg(sc, A_PL_INT_ENABLE0);
933 * sge_timer_cb - perform periodic maintenance of an SGE qset
934 * @data: the SGE queue set to maintain
936 * Runs periodically from a timer to perform maintenance of an SGE queue
937 * set. It performs two tasks:
939 * a) Cleans up any completed Tx descriptors that may still be pending.
940 * Normal descriptor cleanup happens when new packets are added to a Tx
941 * queue so this timer is relatively infrequent and does any cleanup only
942 * if the Tx queue has not seen any new packets in a while. We make a
943 * best effort attempt to reclaim descriptors, in that we don't wait
944 * around if we cannot get a queue's lock (which most likely is because
945 * someone else is queueing new packets and so will also handle the clean
946 * up). Since control queues use immediate data exclusively we don't
947 * bother cleaning them up here.
949 * b) Replenishes Rx queues that have run out due to memory shortage.
950 * Normally new Rx buffers are added when existing ones are consumed but
951 * when out of memory a queue can become empty. We try to add only a few
952 * buffers here, the queue will be replenished fully as these new buffers
953 * are used up if memory shortage has subsided.
955 * c) Return coalesced response queue credits in case a response queue is
958 * d) Ring doorbells for T304 tunnel queues since we have seen doorbell
959 * fifo overflows and the FW doesn't implement any recovery scheme yet.
962 sge_timer_cb(void *arg)
965 if ((sc->flags & USING_MSIX) == 0) {
967 struct port_info *pi;
971 int reclaim_ofl, refill_rx;
973 if (sc->open_device_map == 0)
976 for (i = 0; i < sc->params.nports; i++) {
978 for (j = 0; j < pi->nqsets; j++) {
979 qs = &sc->sge.qs[pi->first_qset + j];
981 reclaim_ofl = txq[TXQ_OFLD].processed - txq[TXQ_OFLD].cleaned;
982 refill_rx = ((qs->fl[0].credits < qs->fl[0].size) ||
983 (qs->fl[1].credits < qs->fl[1].size));
984 if (reclaim_ofl || refill_rx) {
985 taskqueue_enqueue(sc->tq, &pi->timer_reclaim_task);
992 if (sc->params.nports > 2) {
995 for_each_port(sc, i) {
996 struct port_info *pi = &sc->port[i];
998 t3_write_reg(sc, A_SG_KDOORBELL,
1000 (FW_TUNNEL_SGEEC_START + pi->first_qset));
1003 if (((sc->flags & USING_MSIX) == 0 || sc->params.nports > 2) &&
1004 sc->open_device_map != 0)
1005 callout_reset(&sc->sge_timer_ch, TX_RECLAIM_PERIOD, sge_timer_cb, sc);
1009 * This is meant to be a catch-all function to keep sge state private
1014 t3_sge_init_adapter(adapter_t *sc)
1016 callout_init(&sc->sge_timer_ch, CALLOUT_MPSAFE);
1017 callout_reset(&sc->sge_timer_ch, TX_RECLAIM_PERIOD, sge_timer_cb, sc);
1018 TASK_INIT(&sc->slow_intr_task, 0, sge_slow_intr_handler, sc);
1023 t3_sge_reset_adapter(adapter_t *sc)
1025 callout_reset(&sc->sge_timer_ch, TX_RECLAIM_PERIOD, sge_timer_cb, sc);
1030 t3_sge_init_port(struct port_info *pi)
1032 TASK_INIT(&pi->timer_reclaim_task, 0, sge_timer_reclaim, pi);
1037 * refill_rspq - replenish an SGE response queue
1038 * @adapter: the adapter
1039 * @q: the response queue to replenish
1040 * @credits: how many new responses to make available
1042 * Replenishes a response queue by making the supplied number of responses
1045 static __inline void
1046 refill_rspq(adapter_t *sc, const struct sge_rspq *q, u_int credits)
1049 /* mbufs are allocated on demand when a rspq entry is processed. */
1050 t3_write_reg(sc, A_SG_RSPQ_CREDIT_RETURN,
1051 V_RSPQ(q->cntxt_id) | V_CREDITS(credits));
1055 sge_txq_reclaim_handler(void *arg, int ncount)
1057 struct sge_qset *qs = arg;
1060 for (i = 0; i < 3; i++)
1061 reclaim_completed_tx(qs, 16, i);
1065 sge_timer_reclaim(void *arg, int ncount)
1067 struct port_info *pi = arg;
1068 int i, nqsets = pi->nqsets;
1069 adapter_t *sc = pi->adapter;
1070 struct sge_qset *qs;
1073 KASSERT((sc->flags & USING_MSIX) == 0,
1074 ("can't call timer reclaim for msi-x"));
1076 for (i = 0; i < nqsets; i++) {
1077 qs = &sc->sge.qs[pi->first_qset + i];
1079 reclaim_completed_tx(qs, 16, TXQ_OFLD);
1080 lock = (sc->flags & USING_MSIX) ? &qs->rspq.lock :
1081 &sc->sge.qs[0].rspq.lock;
1083 if (mtx_trylock(lock)) {
1084 /* XXX currently assume that we are *NOT* polling */
1085 uint32_t status = t3_read_reg(sc, A_SG_RSPQ_FL_STATUS);
1087 if (qs->fl[0].credits < qs->fl[0].size - 16)
1088 __refill_fl(sc, &qs->fl[0]);
1089 if (qs->fl[1].credits < qs->fl[1].size - 16)
1090 __refill_fl(sc, &qs->fl[1]);
1092 if (status & (1 << qs->rspq.cntxt_id)) {
1093 if (qs->rspq.credits) {
1094 refill_rspq(sc, &qs->rspq, 1);
1096 t3_write_reg(sc, A_SG_RSPQ_FL_STATUS,
1097 1 << qs->rspq.cntxt_id);
1106 * init_qset_cntxt - initialize an SGE queue set context info
1107 * @qs: the queue set
1108 * @id: the queue set id
1110 * Initializes the TIDs and context ids for the queues of a queue set.
1113 init_qset_cntxt(struct sge_qset *qs, u_int id)
1116 qs->rspq.cntxt_id = id;
1117 qs->fl[0].cntxt_id = 2 * id;
1118 qs->fl[1].cntxt_id = 2 * id + 1;
1119 qs->txq[TXQ_ETH].cntxt_id = FW_TUNNEL_SGEEC_START + id;
1120 qs->txq[TXQ_ETH].token = FW_TUNNEL_TID_START + id;
1121 qs->txq[TXQ_OFLD].cntxt_id = FW_OFLD_SGEEC_START + id;
1122 qs->txq[TXQ_CTRL].cntxt_id = FW_CTRL_SGEEC_START + id;
1123 qs->txq[TXQ_CTRL].token = FW_CTRL_TID_START + id;
1125 mbufq_init(&qs->txq[TXQ_ETH].sendq);
1126 mbufq_init(&qs->txq[TXQ_OFLD].sendq);
1127 mbufq_init(&qs->txq[TXQ_CTRL].sendq);
1132 txq_prod(struct sge_txq *txq, unsigned int ndesc, struct txq_state *txqs)
1134 txq->in_use += ndesc;
1136 * XXX we don't handle stopping of queue
1137 * presumably start handles this when we bump against the end
1139 txqs->gen = txq->gen;
1140 txq->unacked += ndesc;
1141 txqs->compl = (txq->unacked & 32) << (S_WR_COMPL - 5);
1143 txqs->pidx = txq->pidx;
1146 if (((txqs->pidx > txq->cidx) &&
1147 (txq->pidx < txqs->pidx) &&
1148 (txq->pidx >= txq->cidx)) ||
1149 ((txqs->pidx < txq->cidx) &&
1150 (txq->pidx >= txq-> cidx)) ||
1151 ((txqs->pidx < txq->cidx) &&
1152 (txq->cidx < txqs->pidx)))
1153 panic("txqs->pidx=%d txq->pidx=%d txq->cidx=%d",
1154 txqs->pidx, txq->pidx, txq->cidx);
1156 if (txq->pidx >= txq->size) {
1157 txq->pidx -= txq->size;
1164 * calc_tx_descs - calculate the number of Tx descriptors for a packet
1165 * @m: the packet mbufs
1166 * @nsegs: the number of segments
1168 * Returns the number of Tx descriptors needed for the given Ethernet
1169 * packet. Ethernet packets require addition of WR and CPL headers.
1171 static __inline unsigned int
1172 calc_tx_descs(const struct mbuf *m, int nsegs)
1176 if (m->m_pkthdr.len <= PIO_LEN)
1179 flits = sgl_len(nsegs) + 2;
1180 if (m->m_pkthdr.csum_flags & CSUM_TSO)
1183 return flits_to_desc(flits);
1187 * make_sgl - populate a scatter/gather list for a packet
1188 * @sgp: the SGL to populate
1189 * @segs: the packet dma segments
1190 * @nsegs: the number of segments
1192 * Generates a scatter/gather list for the buffers that make up a packet
1193 * and returns the SGL size in 8-byte words. The caller must size the SGL
1196 static __inline void
1197 make_sgl(struct sg_ent *sgp, bus_dma_segment_t *segs, int nsegs)
1201 for (idx = 0, i = 0; i < nsegs; i++) {
1203 * firmware doesn't like empty segments
1205 if (segs[i].ds_len == 0)
1210 sgp->len[idx] = htobe32(segs[i].ds_len);
1211 sgp->addr[idx] = htobe64(segs[i].ds_addr);
1222 * check_ring_tx_db - check and potentially ring a Tx queue's doorbell
1223 * @adap: the adapter
1226 * Ring the doorbell if a Tx queue is asleep. There is a natural race,
1227 * where the HW is going to sleep just after we checked, however,
1228 * then the interrupt handler will detect the outstanding TX packet
1229 * and ring the doorbell for us.
1231 * When GTS is disabled we unconditionally ring the doorbell.
1233 static __inline void
1234 check_ring_tx_db(adapter_t *adap, struct sge_txq *q, int mustring)
1237 clear_bit(TXQ_LAST_PKT_DB, &q->flags);
1238 if (test_and_set_bit(TXQ_RUNNING, &q->flags) == 0) {
1239 set_bit(TXQ_LAST_PKT_DB, &q->flags);
1241 T3_TRACE1(adap->tb[q->cntxt_id & 7], "doorbell Tx, cntxt %d",
1244 t3_write_reg(adap, A_SG_KDOORBELL,
1245 F_SELEGRCNTX | V_EGRCNTX(q->cntxt_id));
1248 if (mustring || ++q->db_pending >= 32) {
1249 wmb(); /* write descriptors before telling HW */
1250 t3_write_reg(adap, A_SG_KDOORBELL,
1251 F_SELEGRCNTX | V_EGRCNTX(q->cntxt_id));
1257 static __inline void
1258 wr_gen2(struct tx_desc *d, unsigned int gen)
1260 #if SGE_NUM_GENBITS == 2
1261 d->flit[TX_DESC_FLITS - 1] = htobe64(gen);
1266 * write_wr_hdr_sgl - write a WR header and, optionally, SGL
1267 * @ndesc: number of Tx descriptors spanned by the SGL
1268 * @txd: first Tx descriptor to be written
1269 * @txqs: txq state (generation and producer index)
1270 * @txq: the SGE Tx queue
1272 * @flits: number of flits to the start of the SGL in the first descriptor
1273 * @sgl_flits: the SGL size in flits
1274 * @wr_hi: top 32 bits of WR header based on WR type (big endian)
1275 * @wr_lo: low 32 bits of WR header based on WR type (big endian)
1277 * Write a work request header and an associated SGL. If the SGL is
1278 * small enough to fit into one Tx descriptor it has already been written
1279 * and we just need to write the WR header. Otherwise we distribute the
1280 * SGL across the number of descriptors it spans.
1283 write_wr_hdr_sgl(unsigned int ndesc, struct tx_desc *txd, struct txq_state *txqs,
1284 const struct sge_txq *txq, const struct sg_ent *sgl, unsigned int flits,
1285 unsigned int sgl_flits, unsigned int wr_hi, unsigned int wr_lo)
1288 struct work_request_hdr *wrp = (struct work_request_hdr *)txd;
1289 struct tx_sw_desc *txsd = &txq->sdesc[txqs->pidx];
1291 if (__predict_true(ndesc == 1)) {
1292 set_wr_hdr(wrp, htonl(F_WR_SOP | F_WR_EOP | V_WR_DATATYPE(1) |
1293 V_WR_SGLSFLT(flits)) | wr_hi,
1294 htonl(V_WR_LEN(flits + sgl_flits) | V_WR_GEN(txqs->gen)) |
1297 wr_gen2(txd, txqs->gen);
1300 unsigned int ogen = txqs->gen;
1301 const uint64_t *fp = (const uint64_t *)sgl;
1302 struct work_request_hdr *wp = wrp;
1304 wrp->wrh_hi = htonl(F_WR_SOP | V_WR_DATATYPE(1) |
1305 V_WR_SGLSFLT(flits)) | wr_hi;
1308 unsigned int avail = WR_FLITS - flits;
1310 if (avail > sgl_flits)
1312 memcpy(&txd->flit[flits], fp, avail * sizeof(*fp));
1321 if (++txqs->pidx == txq->size) {
1329 * when the head of the mbuf chain
1330 * is freed all clusters will be freed
1333 wrp = (struct work_request_hdr *)txd;
1334 wrp->wrh_hi = htonl(V_WR_DATATYPE(1) |
1335 V_WR_SGLSFLT(1)) | wr_hi;
1336 wrp->wrh_lo = htonl(V_WR_LEN(min(WR_FLITS,
1338 V_WR_GEN(txqs->gen)) | wr_lo;
1339 wr_gen2(txd, txqs->gen);
1342 wrp->wrh_hi |= htonl(F_WR_EOP);
1344 wp->wrh_lo = htonl(V_WR_LEN(WR_FLITS) | V_WR_GEN(ogen)) | wr_lo;
1345 wr_gen2((struct tx_desc *)wp, ogen);
1349 /* sizeof(*eh) + sizeof(*ip) + sizeof(*tcp) */
1350 #define TCPPKTHDRSIZE (ETHER_HDR_LEN + 20 + 20)
1352 #define GET_VTAG(cntrl, m) \
1354 if ((m)->m_flags & M_VLANTAG) \
1355 cntrl |= F_TXPKT_VLAN_VLD | V_TXPKT_VLAN((m)->m_pkthdr.ether_vtag); \
1359 t3_encap(struct sge_qset *qs, struct mbuf **m)
1363 struct sge_txq *txq;
1364 struct txq_state txqs;
1365 struct port_info *pi;
1366 unsigned int ndesc, flits, cntrl, mlen;
1367 int err, nsegs, tso_info = 0;
1369 struct work_request_hdr *wrp;
1370 struct tx_sw_desc *txsd;
1371 struct sg_ent *sgp, *sgl;
1372 uint32_t wr_hi, wr_lo, sgl_flits;
1373 bus_dma_segment_t segs[TX_MAX_SEGS];
1375 struct tx_desc *txd;
1379 txq = &qs->txq[TXQ_ETH];
1380 txd = &txq->desc[txq->pidx];
1381 txsd = &txq->sdesc[txq->pidx];
1387 mtx_assert(&qs->lock, MA_OWNED);
1388 cntrl = V_TXPKT_INTF(pi->txpkt_intf);
1389 KASSERT(m0->m_flags & M_PKTHDR, ("not packet header\n"));
1391 if (m0->m_nextpkt == NULL && m0->m_next != NULL &&
1392 m0->m_pkthdr.csum_flags & (CSUM_TSO))
1393 tso_info = V_LSO_MSS(m0->m_pkthdr.tso_segsz);
1395 if (m0->m_nextpkt != NULL) {
1396 busdma_map_sg_vec(txq->entry_tag, txsd->map, m0, segs, &nsegs);
1400 if ((err = busdma_map_sg_collapse(txq->entry_tag, txsd->map,
1401 &m0, segs, &nsegs))) {
1403 printf("failed ... err=%d\n", err);
1406 mlen = m0->m_pkthdr.len;
1407 ndesc = calc_tx_descs(m0, nsegs);
1409 txq_prod(txq, ndesc, &txqs);
1411 KASSERT(m0->m_pkthdr.len, ("empty packet nsegs=%d", nsegs));
1414 if (m0->m_nextpkt != NULL) {
1415 struct cpl_tx_pkt_batch *cpl_batch = (struct cpl_tx_pkt_batch *)txd;
1419 panic("trying to coalesce %d packets in to one WR", nsegs);
1420 txq->txq_coalesced += nsegs;
1421 wrp = (struct work_request_hdr *)txd;
1422 flits = nsegs*2 + 1;
1424 for (fidx = 1, i = 0; i < nsegs; i++, fidx += 2) {
1425 struct cpl_tx_pkt_batch_entry *cbe;
1427 uint32_t *hflit = (uint32_t *)&flit;
1428 int cflags = m0->m_pkthdr.csum_flags;
1430 cntrl = V_TXPKT_INTF(pi->txpkt_intf);
1431 GET_VTAG(cntrl, m0);
1432 cntrl |= V_TXPKT_OPCODE(CPL_TX_PKT);
1433 if (__predict_false(!(cflags & CSUM_IP)))
1434 cntrl |= F_TXPKT_IPCSUM_DIS;
1435 if (__predict_false(!(cflags & (CSUM_TCP | CSUM_UDP |
1436 CSUM_UDP_IPV6 | CSUM_TCP_IPV6))))
1437 cntrl |= F_TXPKT_L4CSUM_DIS;
1439 hflit[0] = htonl(cntrl);
1440 hflit[1] = htonl(segs[i].ds_len | 0x80000000);
1441 flit |= htobe64(1 << 24);
1442 cbe = &cpl_batch->pkt_entry[i];
1443 cbe->cntrl = hflit[0];
1444 cbe->len = hflit[1];
1445 cbe->addr = htobe64(segs[i].ds_addr);
1448 wr_hi = htonl(F_WR_SOP | F_WR_EOP | V_WR_DATATYPE(1) |
1449 V_WR_SGLSFLT(flits)) |
1450 htonl(V_WR_OP(FW_WROPCODE_TUNNEL_TX_PKT) | txqs.compl);
1451 wr_lo = htonl(V_WR_LEN(flits) |
1452 V_WR_GEN(txqs.gen)) | htonl(V_WR_TID(txq->token));
1453 set_wr_hdr(wrp, wr_hi, wr_lo);
1455 ETHER_BPF_MTAP(pi->ifp, m0);
1456 wr_gen2(txd, txqs.gen);
1457 check_ring_tx_db(sc, txq, 0);
1459 } else if (tso_info) {
1461 struct cpl_tx_pkt_lso *hdr = (struct cpl_tx_pkt_lso *)txd;
1462 struct ether_header *eh;
1467 GET_VTAG(cntrl, m0);
1468 cntrl |= V_TXPKT_OPCODE(CPL_TX_PKT_LSO);
1469 hdr->cntrl = htonl(cntrl);
1470 hdr->len = htonl(mlen | 0x80000000);
1472 if (__predict_false(mlen < TCPPKTHDRSIZE)) {
1473 printf("mbuf=%p,len=%d,tso_segsz=%d,csum_flags=%b,flags=%#x",
1474 m0, mlen, m0->m_pkthdr.tso_segsz,
1475 (int)m0->m_pkthdr.csum_flags, CSUM_BITS, m0->m_flags);
1476 panic("tx tso packet too small");
1479 /* Make sure that ether, ip, tcp headers are all in m0 */
1480 if (__predict_false(m0->m_len < TCPPKTHDRSIZE)) {
1481 m0 = m_pullup(m0, TCPPKTHDRSIZE);
1482 if (__predict_false(m0 == NULL)) {
1483 /* XXX panic probably an overreaction */
1484 panic("couldn't fit header into mbuf");
1488 eh = mtod(m0, struct ether_header *);
1489 eth_type = eh->ether_type;
1490 if (eth_type == htons(ETHERTYPE_VLAN)) {
1491 struct ether_vlan_header *evh = (void *)eh;
1493 tso_info |= V_LSO_ETH_TYPE(CPL_ETH_II_VLAN);
1495 eth_type = evh->evl_proto;
1497 tso_info |= V_LSO_ETH_TYPE(CPL_ETH_II);
1501 if (eth_type == htons(ETHERTYPE_IP)) {
1502 struct ip *ip = l3hdr;
1504 tso_info |= V_LSO_IPHDR_WORDS(ip->ip_hl);
1505 tcp = (struct tcphdr *)(ip + 1);
1506 } else if (eth_type == htons(ETHERTYPE_IPV6)) {
1507 struct ip6_hdr *ip6 = l3hdr;
1509 KASSERT(ip6->ip6_nxt == IPPROTO_TCP,
1510 ("%s: CSUM_TSO with ip6_nxt %d",
1511 __func__, ip6->ip6_nxt));
1513 tso_info |= F_LSO_IPV6;
1514 tso_info |= V_LSO_IPHDR_WORDS(sizeof(*ip6) >> 2);
1515 tcp = (struct tcphdr *)(ip6 + 1);
1517 panic("%s: CSUM_TSO but neither ip nor ip6", __func__);
1519 tso_info |= V_LSO_TCPHDR_WORDS(tcp->th_off);
1520 hdr->lso_info = htonl(tso_info);
1522 if (__predict_false(mlen <= PIO_LEN)) {
1524 * pkt not undersized but fits in PIO_LEN
1525 * Indicates a TSO bug at the higher levels.
1528 m_copydata(m0, 0, mlen, (caddr_t)&txd->flit[3]);
1529 flits = (mlen + 7) / 8 + 3;
1530 wr_hi = htonl(V_WR_BCNTLFLT(mlen & 7) |
1531 V_WR_OP(FW_WROPCODE_TUNNEL_TX_PKT) |
1532 F_WR_SOP | F_WR_EOP | txqs.compl);
1533 wr_lo = htonl(V_WR_LEN(flits) |
1534 V_WR_GEN(txqs.gen) | V_WR_TID(txq->token));
1535 set_wr_hdr(&hdr->wr, wr_hi, wr_lo);
1537 ETHER_BPF_MTAP(pi->ifp, m0);
1538 wr_gen2(txd, txqs.gen);
1539 check_ring_tx_db(sc, txq, 0);
1545 struct cpl_tx_pkt *cpl = (struct cpl_tx_pkt *)txd;
1547 GET_VTAG(cntrl, m0);
1548 cntrl |= V_TXPKT_OPCODE(CPL_TX_PKT);
1549 if (__predict_false(!(m0->m_pkthdr.csum_flags & CSUM_IP)))
1550 cntrl |= F_TXPKT_IPCSUM_DIS;
1551 if (__predict_false(!(m0->m_pkthdr.csum_flags & (CSUM_TCP |
1552 CSUM_UDP | CSUM_UDP_IPV6 | CSUM_TCP_IPV6))))
1553 cntrl |= F_TXPKT_L4CSUM_DIS;
1554 cpl->cntrl = htonl(cntrl);
1555 cpl->len = htonl(mlen | 0x80000000);
1557 if (mlen <= PIO_LEN) {
1559 m_copydata(m0, 0, mlen, (caddr_t)&txd->flit[2]);
1560 flits = (mlen + 7) / 8 + 2;
1562 wr_hi = htonl(V_WR_BCNTLFLT(mlen & 7) |
1563 V_WR_OP(FW_WROPCODE_TUNNEL_TX_PKT) |
1564 F_WR_SOP | F_WR_EOP | txqs.compl);
1565 wr_lo = htonl(V_WR_LEN(flits) |
1566 V_WR_GEN(txqs.gen) | V_WR_TID(txq->token));
1567 set_wr_hdr(&cpl->wr, wr_hi, wr_lo);
1569 ETHER_BPF_MTAP(pi->ifp, m0);
1570 wr_gen2(txd, txqs.gen);
1571 check_ring_tx_db(sc, txq, 0);
1577 wrp = (struct work_request_hdr *)txd;
1578 sgp = (ndesc == 1) ? (struct sg_ent *)&txd->flit[flits] : sgl;
1579 make_sgl(sgp, segs, nsegs);
1581 sgl_flits = sgl_len(nsegs);
1583 ETHER_BPF_MTAP(pi->ifp, m0);
1585 KASSERT(ndesc <= 4, ("ndesc too large %d", ndesc));
1586 wr_hi = htonl(V_WR_OP(FW_WROPCODE_TUNNEL_TX_PKT) | txqs.compl);
1587 wr_lo = htonl(V_WR_TID(txq->token));
1588 write_wr_hdr_sgl(ndesc, txd, &txqs, txq, sgl, flits,
1589 sgl_flits, wr_hi, wr_lo);
1590 check_ring_tx_db(sc, txq, 0);
1596 cxgb_tx_watchdog(void *arg)
1598 struct sge_qset *qs = arg;
1599 struct sge_txq *txq = &qs->txq[TXQ_ETH];
1601 if (qs->coalescing != 0 &&
1602 (txq->in_use <= cxgb_tx_coalesce_enable_stop) &&
1605 else if (qs->coalescing == 0 &&
1606 (txq->in_use >= cxgb_tx_coalesce_enable_start))
1608 if (TXQ_TRYLOCK(qs)) {
1609 qs->qs_flags |= QS_FLUSHING;
1610 cxgb_start_locked(qs);
1611 qs->qs_flags &= ~QS_FLUSHING;
1614 if (qs->port->ifp->if_drv_flags & IFF_DRV_RUNNING)
1615 callout_reset_on(&txq->txq_watchdog, hz/4, cxgb_tx_watchdog,
1616 qs, txq->txq_watchdog.c_cpu);
1620 cxgb_tx_timeout(void *arg)
1622 struct sge_qset *qs = arg;
1623 struct sge_txq *txq = &qs->txq[TXQ_ETH];
1625 if (qs->coalescing == 0 && (txq->in_use >= (txq->size>>3)))
1627 if (TXQ_TRYLOCK(qs)) {
1628 qs->qs_flags |= QS_TIMEOUT;
1629 cxgb_start_locked(qs);
1630 qs->qs_flags &= ~QS_TIMEOUT;
1636 cxgb_start_locked(struct sge_qset *qs)
1638 struct mbuf *m_head = NULL;
1639 struct sge_txq *txq = &qs->txq[TXQ_ETH];
1640 struct port_info *pi = qs->port;
1641 struct ifnet *ifp = pi->ifp;
1643 if (qs->qs_flags & (QS_FLUSHING|QS_TIMEOUT))
1644 reclaim_completed_tx(qs, 0, TXQ_ETH);
1646 if (!pi->link_config.link_ok) {
1650 TXQ_LOCK_ASSERT(qs);
1651 while (!TXQ_RING_EMPTY(qs) && (ifp->if_drv_flags & IFF_DRV_RUNNING) &&
1652 pi->link_config.link_ok) {
1653 reclaim_completed_tx(qs, cxgb_tx_reclaim_threshold, TXQ_ETH);
1655 if (txq->size - txq->in_use <= TX_MAX_DESC)
1658 if ((m_head = cxgb_dequeue(qs)) == NULL)
1661 * Encapsulation can modify our pointer, and or make it
1662 * NULL on failure. In that event, we can't requeue.
1664 if (t3_encap(qs, &m_head) || m_head == NULL)
1670 if (txq->db_pending)
1671 check_ring_tx_db(pi->adapter, txq, 1);
1673 if (!TXQ_RING_EMPTY(qs) && callout_pending(&txq->txq_timer) == 0 &&
1674 pi->link_config.link_ok)
1675 callout_reset_on(&txq->txq_timer, 1, cxgb_tx_timeout,
1676 qs, txq->txq_timer.c_cpu);
1682 cxgb_transmit_locked(struct ifnet *ifp, struct sge_qset *qs, struct mbuf *m)
1684 struct port_info *pi = qs->port;
1685 struct sge_txq *txq = &qs->txq[TXQ_ETH];
1686 struct buf_ring *br = txq->txq_mr;
1689 avail = txq->size - txq->in_use;
1690 TXQ_LOCK_ASSERT(qs);
1693 * We can only do a direct transmit if the following are true:
1694 * - we aren't coalescing (ring < 3/4 full)
1695 * - the link is up -- checked in caller
1696 * - there are no packets enqueued already
1697 * - there is space in hardware transmit queue
1699 if (check_pkt_coalesce(qs) == 0 &&
1700 !TXQ_RING_NEEDS_ENQUEUE(qs) && avail > TX_MAX_DESC) {
1701 if (t3_encap(qs, &m)) {
1703 (error = drbr_enqueue(ifp, br, m)) != 0)
1706 if (txq->db_pending)
1707 check_ring_tx_db(pi->adapter, txq, 1);
1710 * We've bypassed the buf ring so we need to update
1711 * the stats directly
1713 txq->txq_direct_packets++;
1714 txq->txq_direct_bytes += m->m_pkthdr.len;
1716 } else if ((error = drbr_enqueue(ifp, br, m)) != 0)
1719 reclaim_completed_tx(qs, cxgb_tx_reclaim_threshold, TXQ_ETH);
1720 if (!TXQ_RING_EMPTY(qs) && pi->link_config.link_ok &&
1721 (!check_pkt_coalesce(qs) || (drbr_inuse(ifp, br) >= 7)))
1722 cxgb_start_locked(qs);
1723 else if (!TXQ_RING_EMPTY(qs) && !callout_pending(&txq->txq_timer))
1724 callout_reset_on(&txq->txq_timer, 1, cxgb_tx_timeout,
1725 qs, txq->txq_timer.c_cpu);
1730 cxgb_transmit(struct ifnet *ifp, struct mbuf *m)
1732 struct sge_qset *qs;
1733 struct port_info *pi = ifp->if_softc;
1734 int error, qidx = pi->first_qset;
1736 if ((ifp->if_drv_flags & IFF_DRV_RUNNING) == 0
1737 ||(!pi->link_config.link_ok)) {
1742 if (m->m_flags & M_FLOWID)
1743 qidx = (m->m_pkthdr.flowid % pi->nqsets) + pi->first_qset;
1745 qs = &pi->adapter->sge.qs[qidx];
1747 if (TXQ_TRYLOCK(qs)) {
1749 error = cxgb_transmit_locked(ifp, qs, m);
1752 error = drbr_enqueue(ifp, qs->txq[TXQ_ETH].txq_mr, m);
1757 cxgb_qflush(struct ifnet *ifp)
1760 * flush any enqueued mbufs in the buf_rings
1761 * and in the transmit queues
1768 * write_imm - write a packet into a Tx descriptor as immediate data
1769 * @d: the Tx descriptor to write
1771 * @len: the length of packet data to write as immediate data
1772 * @gen: the generation bit value to write
1774 * Writes a packet as immediate data into a Tx descriptor. The packet
1775 * contains a work request at its beginning. We must write the packet
1776 * carefully so the SGE doesn't read accidentally before it's written in
1779 static __inline void
1780 write_imm(struct tx_desc *d, caddr_t src,
1781 unsigned int len, unsigned int gen)
1783 struct work_request_hdr *from = (struct work_request_hdr *)src;
1784 struct work_request_hdr *to = (struct work_request_hdr *)d;
1785 uint32_t wr_hi, wr_lo;
1787 KASSERT(len <= WR_LEN && len >= sizeof(*from),
1788 ("%s: invalid len %d", __func__, len));
1790 memcpy(&to[1], &from[1], len - sizeof(*from));
1791 wr_hi = from->wrh_hi | htonl(F_WR_SOP | F_WR_EOP |
1792 V_WR_BCNTLFLT(len & 7));
1793 wr_lo = from->wrh_lo | htonl(V_WR_GEN(gen) | V_WR_LEN((len + 7) / 8));
1794 set_wr_hdr(to, wr_hi, wr_lo);
1800 * check_desc_avail - check descriptor availability on a send queue
1801 * @adap: the adapter
1803 * @m: the packet needing the descriptors
1804 * @ndesc: the number of Tx descriptors needed
1805 * @qid: the Tx queue number in its queue set (TXQ_OFLD or TXQ_CTRL)
1807 * Checks if the requested number of Tx descriptors is available on an
1808 * SGE send queue. If the queue is already suspended or not enough
1809 * descriptors are available the packet is queued for later transmission.
1810 * Must be called with the Tx queue locked.
1812 * Returns 0 if enough descriptors are available, 1 if there aren't
1813 * enough descriptors and the packet has been queued, and 2 if the caller
1814 * needs to retry because there weren't enough descriptors at the
1815 * beginning of the call but some freed up in the mean time.
1818 check_desc_avail(adapter_t *adap, struct sge_txq *q,
1819 struct mbuf *m, unsigned int ndesc,
1823 * XXX We currently only use this for checking the control queue
1824 * the control queue is only used for binding qsets which happens
1825 * at init time so we are guaranteed enough descriptors
1827 if (__predict_false(!mbufq_empty(&q->sendq))) {
1828 addq_exit: mbufq_tail(&q->sendq, m);
1831 if (__predict_false(q->size - q->in_use < ndesc)) {
1833 struct sge_qset *qs = txq_to_qset(q, qid);
1835 setbit(&qs->txq_stopped, qid);
1836 if (should_restart_tx(q) &&
1837 test_and_clear_bit(qid, &qs->txq_stopped))
1848 * reclaim_completed_tx_imm - reclaim completed control-queue Tx descs
1849 * @q: the SGE control Tx queue
1851 * This is a variant of reclaim_completed_tx() that is used for Tx queues
1852 * that send only immediate data (presently just the control queues) and
1853 * thus do not have any mbufs
1855 static __inline void
1856 reclaim_completed_tx_imm(struct sge_txq *q)
1858 unsigned int reclaim = q->processed - q->cleaned;
1860 q->in_use -= reclaim;
1861 q->cleaned += reclaim;
1865 * ctrl_xmit - send a packet through an SGE control Tx queue
1866 * @adap: the adapter
1867 * @q: the control queue
1870 * Send a packet through an SGE control Tx queue. Packets sent through
1871 * a control queue must fit entirely as immediate data in a single Tx
1872 * descriptor and have no page fragments.
1875 ctrl_xmit(adapter_t *adap, struct sge_qset *qs, struct mbuf *m)
1878 struct work_request_hdr *wrp = mtod(m, struct work_request_hdr *);
1879 struct sge_txq *q = &qs->txq[TXQ_CTRL];
1881 KASSERT(m->m_len <= WR_LEN, ("%s: bad tx data", __func__));
1883 wrp->wrh_hi |= htonl(F_WR_SOP | F_WR_EOP);
1884 wrp->wrh_lo = htonl(V_WR_TID(q->token));
1887 again: reclaim_completed_tx_imm(q);
1889 ret = check_desc_avail(adap, q, m, 1, TXQ_CTRL);
1890 if (__predict_false(ret)) {
1897 write_imm(&q->desc[q->pidx], m->m_data, m->m_len, q->gen);
1900 if (++q->pidx >= q->size) {
1906 t3_write_reg(adap, A_SG_KDOORBELL,
1907 F_SELEGRCNTX | V_EGRCNTX(q->cntxt_id));
1915 * restart_ctrlq - restart a suspended control queue
1916 * @qs: the queue set cotaining the control queue
1918 * Resumes transmission on a suspended Tx control queue.
1921 restart_ctrlq(void *data, int npending)
1924 struct sge_qset *qs = (struct sge_qset *)data;
1925 struct sge_txq *q = &qs->txq[TXQ_CTRL];
1926 adapter_t *adap = qs->port->adapter;
1929 again: reclaim_completed_tx_imm(q);
1931 while (q->in_use < q->size &&
1932 (m = mbufq_dequeue(&q->sendq)) != NULL) {
1934 write_imm(&q->desc[q->pidx], m->m_data, m->m_len, q->gen);
1937 if (++q->pidx >= q->size) {
1943 if (!mbufq_empty(&q->sendq)) {
1944 setbit(&qs->txq_stopped, TXQ_CTRL);
1946 if (should_restart_tx(q) &&
1947 test_and_clear_bit(TXQ_CTRL, &qs->txq_stopped))
1952 t3_write_reg(adap, A_SG_KDOORBELL,
1953 F_SELEGRCNTX | V_EGRCNTX(q->cntxt_id));
1958 * Send a management message through control queue 0
1961 t3_mgmt_tx(struct adapter *adap, struct mbuf *m)
1963 return ctrl_xmit(adap, &adap->sge.qs[0], m);
1967 * free_qset - free the resources of an SGE queue set
1968 * @sc: the controller owning the queue set
1971 * Release the HW and SW resources associated with an SGE queue set, such
1972 * as HW contexts, packet buffers, and descriptor rings. Traffic to the
1973 * queue set must be quiesced prior to calling this.
1976 t3_free_qset(adapter_t *sc, struct sge_qset *q)
1980 reclaim_completed_tx(q, 0, TXQ_ETH);
1981 if (q->txq[TXQ_ETH].txq_mr != NULL)
1982 buf_ring_free(q->txq[TXQ_ETH].txq_mr, M_DEVBUF);
1983 if (q->txq[TXQ_ETH].txq_ifq != NULL) {
1984 ifq_delete(q->txq[TXQ_ETH].txq_ifq);
1985 free(q->txq[TXQ_ETH].txq_ifq, M_DEVBUF);
1988 for (i = 0; i < SGE_RXQ_PER_SET; ++i) {
1989 if (q->fl[i].desc) {
1990 mtx_lock_spin(&sc->sge.reg_lock);
1991 t3_sge_disable_fl(sc, q->fl[i].cntxt_id);
1992 mtx_unlock_spin(&sc->sge.reg_lock);
1993 bus_dmamap_unload(q->fl[i].desc_tag, q->fl[i].desc_map);
1994 bus_dmamem_free(q->fl[i].desc_tag, q->fl[i].desc,
1996 bus_dma_tag_destroy(q->fl[i].desc_tag);
1997 bus_dma_tag_destroy(q->fl[i].entry_tag);
1999 if (q->fl[i].sdesc) {
2000 free_rx_bufs(sc, &q->fl[i]);
2001 free(q->fl[i].sdesc, M_DEVBUF);
2005 mtx_unlock(&q->lock);
2006 MTX_DESTROY(&q->lock);
2007 for (i = 0; i < SGE_TXQ_PER_SET; i++) {
2008 if (q->txq[i].desc) {
2009 mtx_lock_spin(&sc->sge.reg_lock);
2010 t3_sge_enable_ecntxt(sc, q->txq[i].cntxt_id, 0);
2011 mtx_unlock_spin(&sc->sge.reg_lock);
2012 bus_dmamap_unload(q->txq[i].desc_tag,
2013 q->txq[i].desc_map);
2014 bus_dmamem_free(q->txq[i].desc_tag, q->txq[i].desc,
2015 q->txq[i].desc_map);
2016 bus_dma_tag_destroy(q->txq[i].desc_tag);
2017 bus_dma_tag_destroy(q->txq[i].entry_tag);
2019 if (q->txq[i].sdesc) {
2020 free(q->txq[i].sdesc, M_DEVBUF);
2025 mtx_lock_spin(&sc->sge.reg_lock);
2026 t3_sge_disable_rspcntxt(sc, q->rspq.cntxt_id);
2027 mtx_unlock_spin(&sc->sge.reg_lock);
2029 bus_dmamap_unload(q->rspq.desc_tag, q->rspq.desc_map);
2030 bus_dmamem_free(q->rspq.desc_tag, q->rspq.desc,
2032 bus_dma_tag_destroy(q->rspq.desc_tag);
2033 MTX_DESTROY(&q->rspq.lock);
2036 #if defined(INET6) || defined(INET)
2037 tcp_lro_free(&q->lro.ctrl);
2040 bzero(q, sizeof(*q));
2044 * t3_free_sge_resources - free SGE resources
2045 * @sc: the adapter softc
2047 * Frees resources used by the SGE queue sets.
2050 t3_free_sge_resources(adapter_t *sc, int nqsets)
2054 for (i = 0; i < nqsets; ++i) {
2055 TXQ_LOCK(&sc->sge.qs[i]);
2056 t3_free_qset(sc, &sc->sge.qs[i]);
2061 * t3_sge_start - enable SGE
2062 * @sc: the controller softc
2064 * Enables the SGE for DMAs. This is the last step in starting packet
2068 t3_sge_start(adapter_t *sc)
2070 t3_set_reg_field(sc, A_SG_CONTROL, F_GLOBALENABLE, F_GLOBALENABLE);
2074 * t3_sge_stop - disable SGE operation
2077 * Disables the DMA engine. This can be called in emeregencies (e.g.,
2078 * from error interrupts) or from normal process context. In the latter
2079 * case it also disables any pending queue restart tasklets. Note that
2080 * if it is called in interrupt context it cannot disable the restart
2081 * tasklets as it cannot wait, however the tasklets will have no effect
2082 * since the doorbells are disabled and the driver will call this again
2083 * later from process context, at which time the tasklets will be stopped
2084 * if they are still running.
2087 t3_sge_stop(adapter_t *sc)
2091 t3_set_reg_field(sc, A_SG_CONTROL, F_GLOBALENABLE, 0);
2096 for (nqsets = i = 0; i < (sc)->params.nports; i++)
2097 nqsets += sc->port[i].nqsets;
2103 for (i = 0; i < nqsets; ++i) {
2104 struct sge_qset *qs = &sc->sge.qs[i];
2106 taskqueue_drain(sc->tq, &qs->txq[TXQ_OFLD].qresume_task);
2107 taskqueue_drain(sc->tq, &qs->txq[TXQ_CTRL].qresume_task);
2113 * t3_free_tx_desc - reclaims Tx descriptors and their buffers
2114 * @adapter: the adapter
2115 * @q: the Tx queue to reclaim descriptors from
2116 * @reclaimable: the number of descriptors to reclaim
2117 * @m_vec_size: maximum number of buffers to reclaim
2118 * @desc_reclaimed: returns the number of descriptors reclaimed
2120 * Reclaims Tx descriptors from an SGE Tx queue and frees the associated
2121 * Tx buffers. Called with the Tx queue lock held.
2123 * Returns number of buffers of reclaimed
2126 t3_free_tx_desc(struct sge_qset *qs, int reclaimable, int queue)
2128 struct tx_sw_desc *txsd;
2129 unsigned int cidx, mask;
2130 struct sge_txq *q = &qs->txq[queue];
2133 T3_TRACE2(sc->tb[q->cntxt_id & 7],
2134 "reclaiming %u Tx descriptors at cidx %u", reclaimable, cidx);
2138 txsd = &q->sdesc[cidx];
2140 mtx_assert(&qs->lock, MA_OWNED);
2141 while (reclaimable--) {
2142 prefetch(q->sdesc[(cidx + 1) & mask].m);
2143 prefetch(q->sdesc[(cidx + 2) & mask].m);
2145 if (txsd->m != NULL) {
2146 if (txsd->flags & TX_SW_DESC_MAPPED) {
2147 bus_dmamap_unload(q->entry_tag, txsd->map);
2148 txsd->flags &= ~TX_SW_DESC_MAPPED;
2150 m_freem_list(txsd->m);
2156 if (++cidx == q->size) {
2166 * is_new_response - check if a response is newly written
2167 * @r: the response descriptor
2168 * @q: the response queue
2170 * Returns true if a response descriptor contains a yet unprocessed
2174 is_new_response(const struct rsp_desc *r,
2175 const struct sge_rspq *q)
2177 return (r->intr_gen & F_RSPD_GEN2) == q->gen;
2180 #define RSPD_GTS_MASK (F_RSPD_TXQ0_GTS | F_RSPD_TXQ1_GTS)
2181 #define RSPD_CTRL_MASK (RSPD_GTS_MASK | \
2182 V_RSPD_TXQ0_CR(M_RSPD_TXQ0_CR) | \
2183 V_RSPD_TXQ1_CR(M_RSPD_TXQ1_CR) | \
2184 V_RSPD_TXQ2_CR(M_RSPD_TXQ2_CR))
2186 /* How long to delay the next interrupt in case of memory shortage, in 0.1us. */
2187 #define NOMEM_INTR_DELAY 2500
2191 * write_ofld_wr - write an offload work request
2192 * @adap: the adapter
2193 * @m: the packet to send
2195 * @pidx: index of the first Tx descriptor to write
2196 * @gen: the generation value to use
2197 * @ndesc: number of descriptors the packet will occupy
2199 * Write an offload work request to send the supplied packet. The packet
2200 * data already carry the work request with most fields populated.
2203 write_ofld_wr(adapter_t *adap, struct mbuf *m, struct sge_txq *q,
2204 unsigned int pidx, unsigned int gen, unsigned int ndesc)
2206 unsigned int sgl_flits, flits;
2207 int i, idx, nsegs, wrlen;
2208 struct work_request_hdr *from;
2209 struct sg_ent *sgp, t3sgl[TX_MAX_SEGS / 2 + 1];
2210 struct tx_desc *d = &q->desc[pidx];
2211 struct txq_state txqs;
2212 struct sglist_seg *segs;
2213 struct ofld_hdr *oh = mtod(m, struct ofld_hdr *);
2216 from = (void *)(oh + 1); /* Start of WR within mbuf */
2217 wrlen = m->m_len - sizeof(*oh);
2219 if (!(oh->flags & F_HDR_SGL)) {
2220 write_imm(d, (caddr_t)from, wrlen, gen);
2223 * mbuf with "real" immediate tx data will be enqueue_wr'd by
2224 * t3_push_frames and freed in wr_ack. Others, like those sent
2225 * down by close_conn, t3_send_reset, etc. should be freed here.
2227 if (!(oh->flags & F_HDR_DF))
2232 memcpy(&d->flit[1], &from[1], wrlen - sizeof(*from));
2236 sgp = (ndesc == 1) ? (struct sg_ent *)&d->flit[flits] : t3sgl;
2238 nsegs = sgl->sg_nseg;
2239 segs = sgl->sg_segs;
2240 for (idx = 0, i = 0; i < nsegs; i++) {
2241 KASSERT(segs[i].ss_len, ("%s: 0 len in sgl", __func__));
2244 sgp->len[idx] = htobe32(segs[i].ss_len);
2245 sgp->addr[idx] = htobe64(segs[i].ss_paddr);
2253 sgl_flits = sgl_len(nsegs);
2258 write_wr_hdr_sgl(ndesc, d, &txqs, q, t3sgl, flits, sgl_flits,
2259 from->wrh_hi, from->wrh_lo);
2263 * ofld_xmit - send a packet through an offload queue
2264 * @adap: the adapter
2265 * @q: the Tx offload queue
2268 * Send an offload packet through an SGE offload queue.
2271 ofld_xmit(adapter_t *adap, struct sge_qset *qs, struct mbuf *m)
2275 unsigned int pidx, gen;
2276 struct sge_txq *q = &qs->txq[TXQ_OFLD];
2277 struct ofld_hdr *oh = mtod(m, struct ofld_hdr *);
2279 ndesc = G_HDR_NDESC(oh->flags);
2282 again: reclaim_completed_tx(qs, 16, TXQ_OFLD);
2283 ret = check_desc_avail(adap, q, m, ndesc, TXQ_OFLD);
2284 if (__predict_false(ret)) {
2296 if (q->pidx >= q->size) {
2301 write_ofld_wr(adap, m, q, pidx, gen, ndesc);
2302 check_ring_tx_db(adap, q, 1);
2309 * restart_offloadq - restart a suspended offload queue
2310 * @qs: the queue set cotaining the offload queue
2312 * Resumes transmission on a suspended Tx offload queue.
2315 restart_offloadq(void *data, int npending)
2318 struct sge_qset *qs = data;
2319 struct sge_txq *q = &qs->txq[TXQ_OFLD];
2320 adapter_t *adap = qs->port->adapter;
2324 again: cleaned = reclaim_completed_tx(qs, 16, TXQ_OFLD);
2326 while ((m = mbufq_peek(&q->sendq)) != NULL) {
2327 unsigned int gen, pidx;
2328 struct ofld_hdr *oh = mtod(m, struct ofld_hdr *);
2329 unsigned int ndesc = G_HDR_NDESC(oh->flags);
2331 if (__predict_false(q->size - q->in_use < ndesc)) {
2332 setbit(&qs->txq_stopped, TXQ_OFLD);
2333 if (should_restart_tx(q) &&
2334 test_and_clear_bit(TXQ_OFLD, &qs->txq_stopped))
2344 if (q->pidx >= q->size) {
2349 (void)mbufq_dequeue(&q->sendq);
2351 write_ofld_wr(adap, m, q, pidx, gen, ndesc);
2355 set_bit(TXQ_RUNNING, &q->flags);
2356 set_bit(TXQ_LAST_PKT_DB, &q->flags);
2360 t3_write_reg(adap, A_SG_KDOORBELL,
2361 F_SELEGRCNTX | V_EGRCNTX(q->cntxt_id));
2365 * t3_offload_tx - send an offload packet
2368 * Sends an offload packet. We use the packet priority to select the
2369 * appropriate Tx queue as follows: bit 0 indicates whether the packet
2370 * should be sent as regular or control, bits 1-3 select the queue set.
2373 t3_offload_tx(struct adapter *sc, struct mbuf *m)
2375 struct ofld_hdr *oh = mtod(m, struct ofld_hdr *);
2376 struct sge_qset *qs = &sc->sge.qs[G_HDR_QSET(oh->flags)];
2378 if (oh->flags & F_HDR_CTRL) {
2379 m_adj(m, sizeof (*oh)); /* trim ofld_hdr off */
2380 return (ctrl_xmit(sc, qs, m));
2382 return (ofld_xmit(sc, qs, m));
2387 restart_tx(struct sge_qset *qs)
2389 struct adapter *sc = qs->port->adapter;
2391 if (isset(&qs->txq_stopped, TXQ_OFLD) &&
2392 should_restart_tx(&qs->txq[TXQ_OFLD]) &&
2393 test_and_clear_bit(TXQ_OFLD, &qs->txq_stopped)) {
2394 qs->txq[TXQ_OFLD].restarts++;
2395 taskqueue_enqueue(sc->tq, &qs->txq[TXQ_OFLD].qresume_task);
2398 if (isset(&qs->txq_stopped, TXQ_CTRL) &&
2399 should_restart_tx(&qs->txq[TXQ_CTRL]) &&
2400 test_and_clear_bit(TXQ_CTRL, &qs->txq_stopped)) {
2401 qs->txq[TXQ_CTRL].restarts++;
2402 taskqueue_enqueue(sc->tq, &qs->txq[TXQ_CTRL].qresume_task);
2407 * t3_sge_alloc_qset - initialize an SGE queue set
2408 * @sc: the controller softc
2409 * @id: the queue set id
2410 * @nports: how many Ethernet ports will be using this queue set
2411 * @irq_vec_idx: the IRQ vector index for response queue interrupts
2412 * @p: configuration parameters for this queue set
2413 * @ntxq: number of Tx queues for the queue set
2414 * @pi: port info for queue set
2416 * Allocate resources and initialize an SGE queue set. A queue set
2417 * comprises a response queue, two Rx free-buffer queues, and up to 3
2418 * Tx queues. The Tx queues are assigned roles in the order Ethernet
2419 * queue, offload queue, and control queue.
2422 t3_sge_alloc_qset(adapter_t *sc, u_int id, int nports, int irq_vec_idx,
2423 const struct qset_params *p, int ntxq, struct port_info *pi)
2425 struct sge_qset *q = &sc->sge.qs[id];
2428 MTX_INIT(&q->lock, q->namebuf, NULL, MTX_DEF);
2432 if ((q->txq[TXQ_ETH].txq_mr = buf_ring_alloc(cxgb_txq_buf_ring_size,
2433 M_DEVBUF, M_WAITOK, &q->lock)) == NULL) {
2434 device_printf(sc->dev, "failed to allocate mbuf ring\n");
2437 if ((q->txq[TXQ_ETH].txq_ifq = malloc(sizeof(struct ifaltq), M_DEVBUF,
2438 M_NOWAIT | M_ZERO)) == NULL) {
2439 device_printf(sc->dev, "failed to allocate ifq\n");
2442 ifq_init(q->txq[TXQ_ETH].txq_ifq, pi->ifp);
2443 callout_init(&q->txq[TXQ_ETH].txq_timer, 1);
2444 callout_init(&q->txq[TXQ_ETH].txq_watchdog, 1);
2445 q->txq[TXQ_ETH].txq_timer.c_cpu = id % mp_ncpus;
2446 q->txq[TXQ_ETH].txq_watchdog.c_cpu = id % mp_ncpus;
2448 init_qset_cntxt(q, id);
2450 if ((ret = alloc_ring(sc, p->fl_size, sizeof(struct rx_desc),
2451 sizeof(struct rx_sw_desc), &q->fl[0].phys_addr,
2452 &q->fl[0].desc, &q->fl[0].sdesc,
2453 &q->fl[0].desc_tag, &q->fl[0].desc_map,
2454 sc->rx_dmat, &q->fl[0].entry_tag)) != 0) {
2455 printf("error %d from alloc ring fl0\n", ret);
2459 if ((ret = alloc_ring(sc, p->jumbo_size, sizeof(struct rx_desc),
2460 sizeof(struct rx_sw_desc), &q->fl[1].phys_addr,
2461 &q->fl[1].desc, &q->fl[1].sdesc,
2462 &q->fl[1].desc_tag, &q->fl[1].desc_map,
2463 sc->rx_jumbo_dmat, &q->fl[1].entry_tag)) != 0) {
2464 printf("error %d from alloc ring fl1\n", ret);
2468 if ((ret = alloc_ring(sc, p->rspq_size, sizeof(struct rsp_desc), 0,
2469 &q->rspq.phys_addr, &q->rspq.desc, NULL,
2470 &q->rspq.desc_tag, &q->rspq.desc_map,
2471 NULL, NULL)) != 0) {
2472 printf("error %d from alloc ring rspq\n", ret);
2476 snprintf(q->rspq.lockbuf, RSPQ_NAME_LEN, "t3 rspq lock %d:%d",
2477 device_get_unit(sc->dev), irq_vec_idx);
2478 MTX_INIT(&q->rspq.lock, q->rspq.lockbuf, NULL, MTX_DEF);
2480 for (i = 0; i < ntxq; ++i) {
2481 size_t sz = i == TXQ_CTRL ? 0 : sizeof(struct tx_sw_desc);
2483 if ((ret = alloc_ring(sc, p->txq_size[i],
2484 sizeof(struct tx_desc), sz,
2485 &q->txq[i].phys_addr, &q->txq[i].desc,
2486 &q->txq[i].sdesc, &q->txq[i].desc_tag,
2487 &q->txq[i].desc_map,
2488 sc->tx_dmat, &q->txq[i].entry_tag)) != 0) {
2489 printf("error %d from alloc ring tx %i\n", ret, i);
2492 mbufq_init(&q->txq[i].sendq);
2494 q->txq[i].size = p->txq_size[i];
2498 TASK_INIT(&q->txq[TXQ_OFLD].qresume_task, 0, restart_offloadq, q);
2500 TASK_INIT(&q->txq[TXQ_CTRL].qresume_task, 0, restart_ctrlq, q);
2501 TASK_INIT(&q->txq[TXQ_ETH].qreclaim_task, 0, sge_txq_reclaim_handler, q);
2502 TASK_INIT(&q->txq[TXQ_OFLD].qreclaim_task, 0, sge_txq_reclaim_handler, q);
2504 q->fl[0].gen = q->fl[1].gen = 1;
2505 q->fl[0].size = p->fl_size;
2506 q->fl[1].size = p->jumbo_size;
2510 q->rspq.size = p->rspq_size;
2512 q->txq[TXQ_ETH].stop_thres = nports *
2513 flits_to_desc(sgl_len(TX_MAX_SEGS + 1) + 3);
2515 q->fl[0].buf_size = MCLBYTES;
2516 q->fl[0].zone = zone_pack;
2517 q->fl[0].type = EXT_PACKET;
2519 if (p->jumbo_buf_size == MJUM16BYTES) {
2520 q->fl[1].zone = zone_jumbo16;
2521 q->fl[1].type = EXT_JUMBO16;
2522 } else if (p->jumbo_buf_size == MJUM9BYTES) {
2523 q->fl[1].zone = zone_jumbo9;
2524 q->fl[1].type = EXT_JUMBO9;
2525 } else if (p->jumbo_buf_size == MJUMPAGESIZE) {
2526 q->fl[1].zone = zone_jumbop;
2527 q->fl[1].type = EXT_JUMBOP;
2529 KASSERT(0, ("can't deal with jumbo_buf_size %d.", p->jumbo_buf_size));
2533 q->fl[1].buf_size = p->jumbo_buf_size;
2535 /* Allocate and setup the lro_ctrl structure */
2536 q->lro.enabled = !!(pi->ifp->if_capenable & IFCAP_LRO);
2537 #if defined(INET6) || defined(INET)
2538 ret = tcp_lro_init(&q->lro.ctrl);
2540 printf("error %d from tcp_lro_init\n", ret);
2544 q->lro.ctrl.ifp = pi->ifp;
2546 mtx_lock_spin(&sc->sge.reg_lock);
2547 ret = -t3_sge_init_rspcntxt(sc, q->rspq.cntxt_id, irq_vec_idx,
2548 q->rspq.phys_addr, q->rspq.size,
2549 q->fl[0].buf_size, 1, 0);
2551 printf("error %d from t3_sge_init_rspcntxt\n", ret);
2555 for (i = 0; i < SGE_RXQ_PER_SET; ++i) {
2556 ret = -t3_sge_init_flcntxt(sc, q->fl[i].cntxt_id, 0,
2557 q->fl[i].phys_addr, q->fl[i].size,
2558 q->fl[i].buf_size, p->cong_thres, 1,
2561 printf("error %d from t3_sge_init_flcntxt for index i=%d\n", ret, i);
2566 ret = -t3_sge_init_ecntxt(sc, q->txq[TXQ_ETH].cntxt_id, USE_GTS,
2567 SGE_CNTXT_ETH, id, q->txq[TXQ_ETH].phys_addr,
2568 q->txq[TXQ_ETH].size, q->txq[TXQ_ETH].token,
2571 printf("error %d from t3_sge_init_ecntxt\n", ret);
2576 ret = -t3_sge_init_ecntxt(sc, q->txq[TXQ_OFLD].cntxt_id,
2577 USE_GTS, SGE_CNTXT_OFLD, id,
2578 q->txq[TXQ_OFLD].phys_addr,
2579 q->txq[TXQ_OFLD].size, 0, 1, 0);
2581 printf("error %d from t3_sge_init_ecntxt\n", ret);
2587 ret = -t3_sge_init_ecntxt(sc, q->txq[TXQ_CTRL].cntxt_id, 0,
2589 q->txq[TXQ_CTRL].phys_addr,
2590 q->txq[TXQ_CTRL].size,
2591 q->txq[TXQ_CTRL].token, 1, 0);
2593 printf("error %d from t3_sge_init_ecntxt\n", ret);
2598 mtx_unlock_spin(&sc->sge.reg_lock);
2599 t3_update_qset_coalesce(q, p);
2601 refill_fl(sc, &q->fl[0], q->fl[0].size);
2602 refill_fl(sc, &q->fl[1], q->fl[1].size);
2603 refill_rspq(sc, &q->rspq, q->rspq.size - 1);
2605 t3_write_reg(sc, A_SG_GTS, V_RSPQ(q->rspq.cntxt_id) |
2606 V_NEWTIMER(q->rspq.holdoff_tmr));
2611 mtx_unlock_spin(&sc->sge.reg_lock);
2614 t3_free_qset(sc, q);
2620 * Remove CPL_RX_PKT headers from the mbuf and reduce it to a regular mbuf with
2621 * ethernet data. Hardware assistance with various checksums and any vlan tag
2622 * will also be taken into account here.
2625 t3_rx_eth(struct adapter *adap, struct mbuf *m, int ethpad)
2627 struct cpl_rx_pkt *cpl = (struct cpl_rx_pkt *)(mtod(m, uint8_t *) + ethpad);
2628 struct port_info *pi = &adap->port[adap->rxpkt_map[cpl->iff]];
2629 struct ifnet *ifp = pi->ifp;
2631 if (cpl->vlan_valid) {
2632 m->m_pkthdr.ether_vtag = ntohs(cpl->vlan);
2633 m->m_flags |= M_VLANTAG;
2636 m->m_pkthdr.rcvif = ifp;
2638 * adjust after conversion to mbuf chain
2640 m->m_pkthdr.len -= (sizeof(*cpl) + ethpad);
2641 m->m_len -= (sizeof(*cpl) + ethpad);
2642 m->m_data += (sizeof(*cpl) + ethpad);
2644 if (!cpl->fragment && cpl->csum_valid && cpl->csum == 0xffff) {
2645 struct ether_header *eh = mtod(m, void *);
2648 if (eh->ether_type == htons(ETHERTYPE_VLAN)) {
2649 struct ether_vlan_header *evh = mtod(m, void *);
2651 eh_type = evh->evl_proto;
2653 eh_type = eh->ether_type;
2655 if (ifp->if_capenable & IFCAP_RXCSUM &&
2656 eh_type == htons(ETHERTYPE_IP)) {
2657 m->m_pkthdr.csum_flags = (CSUM_IP_CHECKED |
2658 CSUM_IP_VALID | CSUM_DATA_VALID | CSUM_PSEUDO_HDR);
2659 m->m_pkthdr.csum_data = 0xffff;
2660 } else if (ifp->if_capenable & IFCAP_RXCSUM_IPV6 &&
2661 eh_type == htons(ETHERTYPE_IPV6)) {
2662 m->m_pkthdr.csum_flags = (CSUM_DATA_VALID_IPV6 |
2664 m->m_pkthdr.csum_data = 0xffff;
2670 * get_packet - return the next ingress packet buffer from a free list
2671 * @adap: the adapter that received the packet
2672 * @drop_thres: # of remaining buffers before we start dropping packets
2673 * @qs: the qset that the SGE free list holding the packet belongs to
2674 * @mh: the mbuf header, contains a pointer to the head and tail of the mbuf chain
2675 * @r: response descriptor
2677 * Get the next packet from a free list and complete setup of the
2678 * sk_buff. If the packet is small we make a copy and recycle the
2679 * original buffer, otherwise we use the original buffer itself. If a
2680 * positive drop threshold is supplied packets are dropped and their
2681 * buffers recycled if (a) the number of remaining buffers is under the
2682 * threshold and the packet is too big to copy, or (b) the packet should
2683 * be copied but there is no memory for the copy.
2686 get_packet(adapter_t *adap, unsigned int drop_thres, struct sge_qset *qs,
2687 struct t3_mbuf_hdr *mh, struct rsp_desc *r)
2690 unsigned int len_cq = ntohl(r->len_cq);
2691 struct sge_fl *fl = (len_cq & F_RSPD_FLQ) ? &qs->fl[1] : &qs->fl[0];
2692 int mask, cidx = fl->cidx;
2693 struct rx_sw_desc *sd = &fl->sdesc[cidx];
2694 uint32_t len = G_RSPD_LEN(len_cq);
2695 uint32_t flags = M_EXT;
2696 uint8_t sopeop = G_RSPD_SOP_EOP(ntohl(r->flags));
2701 mask = fl->size - 1;
2702 prefetch(fl->sdesc[(cidx + 1) & mask].m);
2703 prefetch(fl->sdesc[(cidx + 2) & mask].m);
2704 prefetch(fl->sdesc[(cidx + 1) & mask].rxsd_cl);
2705 prefetch(fl->sdesc[(cidx + 2) & mask].rxsd_cl);
2708 bus_dmamap_sync(fl->entry_tag, sd->map, BUS_DMASYNC_POSTREAD);
2710 if (recycle_enable && len <= SGE_RX_COPY_THRES &&
2711 sopeop == RSPQ_SOP_EOP) {
2712 if ((m = m_gethdr(M_NOWAIT, MT_DATA)) == NULL)
2714 cl = mtod(m, void *);
2715 memcpy(cl, sd->rxsd_cl, len);
2716 recycle_rx_buf(adap, fl, fl->cidx);
2717 m->m_pkthdr.len = m->m_len = len;
2719 mh->mh_head = mh->mh_tail = m;
2724 bus_dmamap_unload(fl->entry_tag, sd->map);
2728 if ((sopeop == RSPQ_SOP_EOP) ||
2729 (sopeop == RSPQ_SOP))
2731 m_init(m, fl->zone, fl->buf_size, M_NOWAIT, MT_DATA, flags);
2732 if (fl->zone == zone_pack) {
2734 * restore clobbered data pointer
2736 m->m_data = m->m_ext.ext_buf;
2738 m_cljset(m, cl, fl->type);
2747 mh->mh_head = mh->mh_tail = m;
2748 m->m_pkthdr.len = len;
2753 case RSPQ_NSOP_NEOP:
2754 if (mh->mh_tail == NULL) {
2755 log(LOG_ERR, "discarding intermediate descriptor entry\n");
2759 mh->mh_tail->m_next = m;
2761 mh->mh_head->m_pkthdr.len += len;
2765 printf("len=%d pktlen=%d\n", m->m_len, m->m_pkthdr.len);
2767 if (++fl->cidx == fl->size)
2774 * handle_rsp_cntrl_info - handles control information in a response
2775 * @qs: the queue set corresponding to the response
2776 * @flags: the response control flags
2778 * Handles the control information of an SGE response, such as GTS
2779 * indications and completion credits for the queue set's Tx queues.
2780 * HW coalesces credits, we don't do any extra SW coalescing.
2782 static __inline void
2783 handle_rsp_cntrl_info(struct sge_qset *qs, uint32_t flags)
2785 unsigned int credits;
2788 if (flags & F_RSPD_TXQ0_GTS)
2789 clear_bit(TXQ_RUNNING, &qs->txq[TXQ_ETH].flags);
2791 credits = G_RSPD_TXQ0_CR(flags);
2793 qs->txq[TXQ_ETH].processed += credits;
2795 credits = G_RSPD_TXQ2_CR(flags);
2797 qs->txq[TXQ_CTRL].processed += credits;
2800 if (flags & F_RSPD_TXQ1_GTS)
2801 clear_bit(TXQ_RUNNING, &qs->txq[TXQ_OFLD].flags);
2803 credits = G_RSPD_TXQ1_CR(flags);
2805 qs->txq[TXQ_OFLD].processed += credits;
2810 check_ring_db(adapter_t *adap, struct sge_qset *qs,
2811 unsigned int sleeping)
2817 * process_responses - process responses from an SGE response queue
2818 * @adap: the adapter
2819 * @qs: the queue set to which the response queue belongs
2820 * @budget: how many responses can be processed in this round
2822 * Process responses from an SGE response queue up to the supplied budget.
2823 * Responses include received packets as well as credits and other events
2824 * for the queues that belong to the response queue's queue set.
2825 * A negative budget is effectively unlimited.
2827 * Additionally choose the interrupt holdoff time for the next interrupt
2828 * on this queue. If the system is under memory shortage use a fairly
2829 * long delay to help recovery.
2832 process_responses(adapter_t *adap, struct sge_qset *qs, int budget)
2834 struct sge_rspq *rspq = &qs->rspq;
2835 struct rsp_desc *r = &rspq->desc[rspq->cidx];
2836 int budget_left = budget;
2837 unsigned int sleeping = 0;
2838 #if defined(INET6) || defined(INET)
2839 int lro_enabled = qs->lro.enabled;
2841 struct lro_ctrl *lro_ctrl = &qs->lro.ctrl;
2843 struct t3_mbuf_hdr *mh = &rspq->rspq_mh;
2845 static int last_holdoff = 0;
2846 if (cxgb_debug && rspq->holdoff_tmr != last_holdoff) {
2847 printf("next_holdoff=%d\n", rspq->holdoff_tmr);
2848 last_holdoff = rspq->holdoff_tmr;
2851 rspq->next_holdoff = rspq->holdoff_tmr;
2853 while (__predict_true(budget_left && is_new_response(r, rspq))) {
2854 int eth, eop = 0, ethpad = 0;
2855 uint32_t flags = ntohl(r->flags);
2856 uint32_t rss_hash = be32toh(r->rss_hdr.rss_hash_val);
2857 uint8_t opcode = r->rss_hdr.opcode;
2859 eth = (opcode == CPL_RX_PKT);
2861 if (__predict_false(flags & F_RSPD_ASYNC_NOTIF)) {
2865 printf("async notification\n");
2867 if (mh->mh_head == NULL) {
2868 mh->mh_head = m_gethdr(M_NOWAIT, MT_DATA);
2871 m = m_gethdr(M_NOWAIT, MT_DATA);
2876 memcpy(mtod(m, char *), r, AN_PKT_SIZE);
2877 m->m_len = m->m_pkthdr.len = AN_PKT_SIZE;
2878 *mtod(m, char *) = CPL_ASYNC_NOTIF;
2879 opcode = CPL_ASYNC_NOTIF;
2881 rspq->async_notif++;
2883 } else if (flags & F_RSPD_IMM_DATA_VALID) {
2884 struct mbuf *m = m_gethdr(M_NOWAIT, MT_DATA);
2888 rspq->next_holdoff = NOMEM_INTR_DELAY;
2892 if (mh->mh_head == NULL)
2895 mh->mh_tail->m_next = m;
2898 get_imm_packet(adap, r, m);
2899 mh->mh_head->m_pkthdr.len += m->m_len;
2902 } else if (r->len_cq) {
2903 int drop_thresh = eth ? SGE_RX_DROP_THRES : 0;
2905 eop = get_packet(adap, drop_thresh, qs, mh, r);
2907 if (r->rss_hdr.hash_type && !adap->timestamp)
2908 mh->mh_head->m_flags |= M_FLOWID;
2909 mh->mh_head->m_pkthdr.flowid = rss_hash;
2917 if (flags & RSPD_CTRL_MASK) {
2918 sleeping |= flags & RSPD_GTS_MASK;
2919 handle_rsp_cntrl_info(qs, flags);
2923 rspq->offload_pkts++;
2925 adap->cpl_handler[opcode](qs, r, mh->mh_head);
2927 m_freem(mh->mh_head);
2930 } else if (eth && eop) {
2931 struct mbuf *m = mh->mh_head;
2933 t3_rx_eth(adap, m, ethpad);
2936 * The T304 sends incoming packets on any qset. If LRO
2937 * is also enabled, we could end up sending packet up
2938 * lro_ctrl->ifp's input. That is incorrect.
2940 * The mbuf's rcvif was derived from the cpl header and
2941 * is accurate. Skip LRO and just use that.
2943 #if defined(INET6) || defined(INET)
2944 skip_lro = __predict_false(qs->port->ifp != m->m_pkthdr.rcvif);
2946 if (lro_enabled && lro_ctrl->lro_cnt && !skip_lro
2947 && (tcp_lro_rx(lro_ctrl, m, 0) == 0)
2949 /* successfully queue'd for LRO */
2954 * LRO not enabled, packet unsuitable for LRO,
2955 * or unable to queue. Pass it up right now in
2958 struct ifnet *ifp = m->m_pkthdr.rcvif;
2959 (*ifp->if_input)(ifp, m);
2966 if (__predict_false(++rspq->cidx == rspq->size)) {
2972 if (++rspq->credits >= 64) {
2973 refill_rspq(adap, rspq, rspq->credits);
2976 __refill_fl_lt(adap, &qs->fl[0], 32);
2977 __refill_fl_lt(adap, &qs->fl[1], 32);
2981 #if defined(INET6) || defined(INET)
2983 while (!SLIST_EMPTY(&lro_ctrl->lro_active)) {
2984 struct lro_entry *queued = SLIST_FIRST(&lro_ctrl->lro_active);
2985 SLIST_REMOVE_HEAD(&lro_ctrl->lro_active, next);
2986 tcp_lro_flush(lro_ctrl, queued);
2991 check_ring_db(adap, qs, sleeping);
2993 mb(); /* commit Tx queue processed updates */
2994 if (__predict_false(qs->txq_stopped > 1))
2997 __refill_fl_lt(adap, &qs->fl[0], 512);
2998 __refill_fl_lt(adap, &qs->fl[1], 512);
2999 budget -= budget_left;
3004 * A helper function that processes responses and issues GTS.
3007 process_responses_gts(adapter_t *adap, struct sge_rspq *rq)
3010 static int last_holdoff = 0;
3012 work = process_responses(adap, rspq_to_qset(rq), -1);
3014 if (cxgb_debug && (rq->next_holdoff != last_holdoff)) {
3015 printf("next_holdoff=%d\n", rq->next_holdoff);
3016 last_holdoff = rq->next_holdoff;
3018 t3_write_reg(adap, A_SG_GTS, V_RSPQ(rq->cntxt_id) |
3019 V_NEWTIMER(rq->next_holdoff) | V_NEWINDEX(rq->cidx));
3026 * Interrupt handler for legacy INTx interrupts for T3B-based cards.
3027 * Handles data events from SGE response queues as well as error and other
3028 * async events as they all use the same interrupt pin. We use one SGE
3029 * response queue per port in this mode and protect all response queues with
3033 t3b_intr(void *data)
3036 adapter_t *adap = data;
3037 struct sge_rspq *q0 = &adap->sge.qs[0].rspq;
3039 t3_write_reg(adap, A_PL_CLI, 0);
3040 map = t3_read_reg(adap, A_SG_DATA_INTR);
3045 if (__predict_false(map & F_ERRINTR)) {
3046 t3_write_reg(adap, A_PL_INT_ENABLE0, 0);
3047 (void) t3_read_reg(adap, A_PL_INT_ENABLE0);
3048 taskqueue_enqueue(adap->tq, &adap->slow_intr_task);
3051 mtx_lock(&q0->lock);
3052 for_each_port(adap, i)
3054 process_responses_gts(adap, &adap->sge.qs[i].rspq);
3055 mtx_unlock(&q0->lock);
3059 * The MSI interrupt handler. This needs to handle data events from SGE
3060 * response queues as well as error and other async events as they all use
3061 * the same MSI vector. We use one SGE response queue per port in this mode
3062 * and protect all response queues with queue 0's lock.
3065 t3_intr_msi(void *data)
3067 adapter_t *adap = data;
3068 struct sge_rspq *q0 = &adap->sge.qs[0].rspq;
3069 int i, new_packets = 0;
3071 mtx_lock(&q0->lock);
3073 for_each_port(adap, i)
3074 if (process_responses_gts(adap, &adap->sge.qs[i].rspq))
3076 mtx_unlock(&q0->lock);
3077 if (new_packets == 0) {
3078 t3_write_reg(adap, A_PL_INT_ENABLE0, 0);
3079 (void) t3_read_reg(adap, A_PL_INT_ENABLE0);
3080 taskqueue_enqueue(adap->tq, &adap->slow_intr_task);
3085 t3_intr_msix(void *data)
3087 struct sge_qset *qs = data;
3088 adapter_t *adap = qs->port->adapter;
3089 struct sge_rspq *rspq = &qs->rspq;
3091 if (process_responses_gts(adap, rspq) == 0)
3092 rspq->unhandled_irqs++;
3095 #define QDUMP_SBUF_SIZE 32 * 400
3097 t3_dump_rspq(SYSCTL_HANDLER_ARGS)
3099 struct sge_rspq *rspq;
3100 struct sge_qset *qs;
3101 int i, err, dump_end, idx;
3103 struct rsp_desc *rspd;
3107 qs = rspq_to_qset(rspq);
3108 if (rspq->rspq_dump_count == 0)
3110 if (rspq->rspq_dump_count > RSPQ_Q_SIZE) {
3112 "dump count is too large %d\n", rspq->rspq_dump_count);
3113 rspq->rspq_dump_count = 0;
3116 if (rspq->rspq_dump_start > (RSPQ_Q_SIZE-1)) {
3118 "dump start of %d is greater than queue size\n",
3119 rspq->rspq_dump_start);
3120 rspq->rspq_dump_start = 0;
3123 err = t3_sge_read_rspq(qs->port->adapter, rspq->cntxt_id, data);
3126 err = sysctl_wire_old_buffer(req, 0);
3129 sb = sbuf_new_for_sysctl(NULL, NULL, QDUMP_SBUF_SIZE, req);
3131 sbuf_printf(sb, " \n index=%u size=%u MSI-X/RspQ=%u intr enable=%u intr armed=%u\n",
3132 (data[0] & 0xffff), data[0] >> 16, ((data[2] >> 20) & 0x3f),
3133 ((data[2] >> 26) & 1), ((data[2] >> 27) & 1));
3134 sbuf_printf(sb, " generation=%u CQ mode=%u FL threshold=%u\n",
3135 ((data[2] >> 28) & 1), ((data[2] >> 31) & 1), data[3]);
3137 sbuf_printf(sb, " start=%d -> end=%d\n", rspq->rspq_dump_start,
3138 (rspq->rspq_dump_start + rspq->rspq_dump_count) & (RSPQ_Q_SIZE-1));
3140 dump_end = rspq->rspq_dump_start + rspq->rspq_dump_count;
3141 for (i = rspq->rspq_dump_start; i < dump_end; i++) {
3142 idx = i & (RSPQ_Q_SIZE-1);
3144 rspd = &rspq->desc[idx];
3145 sbuf_printf(sb, "\tidx=%04d opcode=%02x cpu_idx=%x hash_type=%x cq_idx=%x\n",
3146 idx, rspd->rss_hdr.opcode, rspd->rss_hdr.cpu_idx,
3147 rspd->rss_hdr.hash_type, be16toh(rspd->rss_hdr.cq_idx));
3148 sbuf_printf(sb, "\trss_hash_val=%x flags=%08x len_cq=%x intr_gen=%x\n",
3149 rspd->rss_hdr.rss_hash_val, be32toh(rspd->flags),
3150 be32toh(rspd->len_cq), rspd->intr_gen);
3153 err = sbuf_finish(sb);
3154 /* Output a trailing NUL. */
3156 err = SYSCTL_OUT(req, "", 1);
3162 t3_dump_txq_eth(SYSCTL_HANDLER_ARGS)
3164 struct sge_txq *txq;
3165 struct sge_qset *qs;
3166 int i, j, err, dump_end;
3168 struct tx_desc *txd;
3169 uint32_t *WR, wr_hi, wr_lo, gen;
3173 qs = txq_to_qset(txq, TXQ_ETH);
3174 if (txq->txq_dump_count == 0) {
3177 if (txq->txq_dump_count > TX_ETH_Q_SIZE) {
3179 "dump count is too large %d\n", txq->txq_dump_count);
3180 txq->txq_dump_count = 1;
3183 if (txq->txq_dump_start > (TX_ETH_Q_SIZE-1)) {
3185 "dump start of %d is greater than queue size\n",
3186 txq->txq_dump_start);
3187 txq->txq_dump_start = 0;
3190 err = t3_sge_read_ecntxt(qs->port->adapter, qs->rspq.cntxt_id, data);
3193 err = sysctl_wire_old_buffer(req, 0);
3196 sb = sbuf_new_for_sysctl(NULL, NULL, QDUMP_SBUF_SIZE, req);
3198 sbuf_printf(sb, " \n credits=%u GTS=%u index=%u size=%u rspq#=%u cmdq#=%u\n",
3199 (data[0] & 0x7fff), ((data[0] >> 15) & 1), (data[0] >> 16),
3200 (data[1] & 0xffff), ((data[3] >> 4) & 7), ((data[3] >> 7) & 1));
3201 sbuf_printf(sb, " TUN=%u TOE=%u generation%u uP token=%u valid=%u\n",
3202 ((data[3] >> 8) & 1), ((data[3] >> 9) & 1), ((data[3] >> 10) & 1),
3203 ((data[3] >> 11) & 0xfffff), ((data[3] >> 31) & 1));
3204 sbuf_printf(sb, " qid=%d start=%d -> end=%d\n", qs->idx,
3205 txq->txq_dump_start,
3206 (txq->txq_dump_start + txq->txq_dump_count) & (TX_ETH_Q_SIZE-1));
3208 dump_end = txq->txq_dump_start + txq->txq_dump_count;
3209 for (i = txq->txq_dump_start; i < dump_end; i++) {
3210 txd = &txq->desc[i & (TX_ETH_Q_SIZE-1)];
3211 WR = (uint32_t *)txd->flit;
3212 wr_hi = ntohl(WR[0]);
3213 wr_lo = ntohl(WR[1]);
3214 gen = G_WR_GEN(wr_lo);
3216 sbuf_printf(sb," wr_hi %08x wr_lo %08x gen %d\n",
3218 for (j = 2; j < 30; j += 4)
3219 sbuf_printf(sb, "\t%08x %08x %08x %08x \n",
3220 WR[j], WR[j + 1], WR[j + 2], WR[j + 3]);
3223 err = sbuf_finish(sb);
3224 /* Output a trailing NUL. */
3226 err = SYSCTL_OUT(req, "", 1);
3232 t3_dump_txq_ctrl(SYSCTL_HANDLER_ARGS)
3234 struct sge_txq *txq;
3235 struct sge_qset *qs;
3236 int i, j, err, dump_end;
3238 struct tx_desc *txd;
3239 uint32_t *WR, wr_hi, wr_lo, gen;
3242 qs = txq_to_qset(txq, TXQ_CTRL);
3243 if (txq->txq_dump_count == 0) {
3246 if (txq->txq_dump_count > 256) {
3248 "dump count is too large %d\n", txq->txq_dump_count);
3249 txq->txq_dump_count = 1;
3252 if (txq->txq_dump_start > 255) {
3254 "dump start of %d is greater than queue size\n",
3255 txq->txq_dump_start);
3256 txq->txq_dump_start = 0;
3260 err = sysctl_wire_old_buffer(req, 0);
3263 sb = sbuf_new_for_sysctl(NULL, NULL, QDUMP_SBUF_SIZE, req);
3264 sbuf_printf(sb, " qid=%d start=%d -> end=%d\n", qs->idx,
3265 txq->txq_dump_start,
3266 (txq->txq_dump_start + txq->txq_dump_count) & 255);
3268 dump_end = txq->txq_dump_start + txq->txq_dump_count;
3269 for (i = txq->txq_dump_start; i < dump_end; i++) {
3270 txd = &txq->desc[i & (255)];
3271 WR = (uint32_t *)txd->flit;
3272 wr_hi = ntohl(WR[0]);
3273 wr_lo = ntohl(WR[1]);
3274 gen = G_WR_GEN(wr_lo);
3276 sbuf_printf(sb," wr_hi %08x wr_lo %08x gen %d\n",
3278 for (j = 2; j < 30; j += 4)
3279 sbuf_printf(sb, "\t%08x %08x %08x %08x \n",
3280 WR[j], WR[j + 1], WR[j + 2], WR[j + 3]);
3283 err = sbuf_finish(sb);
3284 /* Output a trailing NUL. */
3286 err = SYSCTL_OUT(req, "", 1);
3292 t3_set_coalesce_usecs(SYSCTL_HANDLER_ARGS)
3294 adapter_t *sc = arg1;
3295 struct qset_params *qsp = &sc->params.sge.qset[0];
3297 struct sge_qset *qs;
3298 int i, j, err, nqsets = 0;
3301 if ((sc->flags & FULL_INIT_DONE) == 0)
3304 coalesce_usecs = qsp->coalesce_usecs;
3305 err = sysctl_handle_int(oidp, &coalesce_usecs, arg2, req);
3310 if (coalesce_usecs == qsp->coalesce_usecs)
3313 for (i = 0; i < sc->params.nports; i++)
3314 for (j = 0; j < sc->port[i].nqsets; j++)
3317 coalesce_usecs = max(1, coalesce_usecs);
3319 for (i = 0; i < nqsets; i++) {
3320 qs = &sc->sge.qs[i];
3321 qsp = &sc->params.sge.qset[i];
3322 qsp->coalesce_usecs = coalesce_usecs;
3324 lock = (sc->flags & USING_MSIX) ? &qs->rspq.lock :
3325 &sc->sge.qs[0].rspq.lock;
3328 t3_update_qset_coalesce(qs, qsp);
3329 t3_write_reg(sc, A_SG_GTS, V_RSPQ(qs->rspq.cntxt_id) |
3330 V_NEWTIMER(qs->rspq.holdoff_tmr));
3338 t3_pkt_timestamp(SYSCTL_HANDLER_ARGS)
3340 adapter_t *sc = arg1;
3343 if ((sc->flags & FULL_INIT_DONE) == 0)
3346 timestamp = sc->timestamp;
3347 rc = sysctl_handle_int(oidp, ×tamp, arg2, req);
3352 if (timestamp != sc->timestamp) {
3353 t3_set_reg_field(sc, A_TP_PC_CONFIG2, F_ENABLERXPKTTMSTPRSS,
3354 timestamp ? F_ENABLERXPKTTMSTPRSS : 0);
3355 sc->timestamp = timestamp;
3362 t3_add_attach_sysctls(adapter_t *sc)
3364 struct sysctl_ctx_list *ctx;
3365 struct sysctl_oid_list *children;
3367 ctx = device_get_sysctl_ctx(sc->dev);
3368 children = SYSCTL_CHILDREN(device_get_sysctl_tree(sc->dev));
3370 /* random information */
3371 SYSCTL_ADD_STRING(ctx, children, OID_AUTO,
3373 CTLFLAG_RD, &sc->fw_version,
3374 0, "firmware version");
3375 SYSCTL_ADD_UINT(ctx, children, OID_AUTO,
3377 CTLFLAG_RD, &sc->params.rev,
3379 SYSCTL_ADD_STRING(ctx, children, OID_AUTO,
3381 CTLFLAG_RD, &sc->port_types,
3382 0, "type of ports");
3383 SYSCTL_ADD_INT(ctx, children, OID_AUTO,
3385 CTLFLAG_RW, &cxgb_debug,
3386 0, "enable verbose debugging output");
3387 SYSCTL_ADD_UQUAD(ctx, children, OID_AUTO, "tunq_coalesce",
3388 CTLFLAG_RD, &sc->tunq_coalesce,
3389 "#tunneled packets freed");
3390 SYSCTL_ADD_INT(ctx, children, OID_AUTO,
3392 CTLFLAG_RD, &txq_fills,
3393 0, "#times txq overrun");
3394 SYSCTL_ADD_UINT(ctx, children, OID_AUTO,
3396 CTLFLAG_RD, &sc->params.vpd.cclk,
3397 0, "core clock frequency (in KHz)");
3401 static const char *rspq_name = "rspq";
3402 static const char *txq_names[] =
3410 sysctl_handle_macstat(SYSCTL_HANDLER_ARGS)
3412 struct port_info *p = arg1;
3418 parg = (uint64_t *) ((uint8_t *)&p->mac.stats + arg2);
3420 t3_mac_update_stats(&p->mac);
3423 return (sysctl_handle_64(oidp, parg, 0, req));
3427 t3_add_configured_sysctls(adapter_t *sc)
3429 struct sysctl_ctx_list *ctx;
3430 struct sysctl_oid_list *children;
3433 ctx = device_get_sysctl_ctx(sc->dev);
3434 children = SYSCTL_CHILDREN(device_get_sysctl_tree(sc->dev));
3436 SYSCTL_ADD_PROC(ctx, children, OID_AUTO,
3438 CTLTYPE_INT|CTLFLAG_RW, sc,
3439 0, t3_set_coalesce_usecs,
3440 "I", "interrupt coalescing timer (us)");
3442 SYSCTL_ADD_PROC(ctx, children, OID_AUTO,
3444 CTLTYPE_INT | CTLFLAG_RW, sc,
3445 0, t3_pkt_timestamp,
3446 "I", "provide packet timestamp instead of connection hash");
3448 for (i = 0; i < sc->params.nports; i++) {
3449 struct port_info *pi = &sc->port[i];
3450 struct sysctl_oid *poid;
3451 struct sysctl_oid_list *poidlist;
3452 struct mac_stats *mstats = &pi->mac.stats;
3454 snprintf(pi->namebuf, PORT_NAME_LEN, "port%d", i);
3455 poid = SYSCTL_ADD_NODE(ctx, children, OID_AUTO,
3456 pi->namebuf, CTLFLAG_RD, NULL, "port statistics");
3457 poidlist = SYSCTL_CHILDREN(poid);
3458 SYSCTL_ADD_UINT(ctx, poidlist, OID_AUTO,
3459 "nqsets", CTLFLAG_RD, &pi->nqsets,
3462 for (j = 0; j < pi->nqsets; j++) {
3463 struct sge_qset *qs = &sc->sge.qs[pi->first_qset + j];
3464 struct sysctl_oid *qspoid, *rspqpoid, *txqpoid,
3465 *ctrlqpoid, *lropoid;
3466 struct sysctl_oid_list *qspoidlist, *rspqpoidlist,
3467 *txqpoidlist, *ctrlqpoidlist,
3469 struct sge_txq *txq = &qs->txq[TXQ_ETH];
3471 snprintf(qs->namebuf, QS_NAME_LEN, "qs%d", j);
3473 qspoid = SYSCTL_ADD_NODE(ctx, poidlist, OID_AUTO,
3474 qs->namebuf, CTLFLAG_RD, NULL, "qset statistics");
3475 qspoidlist = SYSCTL_CHILDREN(qspoid);
3477 SYSCTL_ADD_UINT(ctx, qspoidlist, OID_AUTO, "fl0_empty",
3478 CTLFLAG_RD, &qs->fl[0].empty, 0,
3479 "freelist #0 empty");
3480 SYSCTL_ADD_UINT(ctx, qspoidlist, OID_AUTO, "fl1_empty",
3481 CTLFLAG_RD, &qs->fl[1].empty, 0,
3482 "freelist #1 empty");
3484 rspqpoid = SYSCTL_ADD_NODE(ctx, qspoidlist, OID_AUTO,
3485 rspq_name, CTLFLAG_RD, NULL, "rspq statistics");
3486 rspqpoidlist = SYSCTL_CHILDREN(rspqpoid);
3488 txqpoid = SYSCTL_ADD_NODE(ctx, qspoidlist, OID_AUTO,
3489 txq_names[0], CTLFLAG_RD, NULL, "txq statistics");
3490 txqpoidlist = SYSCTL_CHILDREN(txqpoid);
3492 ctrlqpoid = SYSCTL_ADD_NODE(ctx, qspoidlist, OID_AUTO,
3493 txq_names[2], CTLFLAG_RD, NULL, "ctrlq statistics");
3494 ctrlqpoidlist = SYSCTL_CHILDREN(ctrlqpoid);
3496 lropoid = SYSCTL_ADD_NODE(ctx, qspoidlist, OID_AUTO,
3497 "lro_stats", CTLFLAG_RD, NULL, "LRO statistics");
3498 lropoidlist = SYSCTL_CHILDREN(lropoid);
3500 SYSCTL_ADD_UINT(ctx, rspqpoidlist, OID_AUTO, "size",
3501 CTLFLAG_RD, &qs->rspq.size,
3502 0, "#entries in response queue");
3503 SYSCTL_ADD_UINT(ctx, rspqpoidlist, OID_AUTO, "cidx",
3504 CTLFLAG_RD, &qs->rspq.cidx,
3505 0, "consumer index");
3506 SYSCTL_ADD_UINT(ctx, rspqpoidlist, OID_AUTO, "credits",
3507 CTLFLAG_RD, &qs->rspq.credits,
3509 SYSCTL_ADD_UINT(ctx, rspqpoidlist, OID_AUTO, "starved",
3510 CTLFLAG_RD, &qs->rspq.starved,
3511 0, "#times starved");
3512 SYSCTL_ADD_ULONG(ctx, rspqpoidlist, OID_AUTO, "phys_addr",
3513 CTLFLAG_RD, &qs->rspq.phys_addr,
3514 "physical_address_of the queue");
3515 SYSCTL_ADD_UINT(ctx, rspqpoidlist, OID_AUTO, "dump_start",
3516 CTLFLAG_RW, &qs->rspq.rspq_dump_start,
3517 0, "start rspq dump entry");
3518 SYSCTL_ADD_UINT(ctx, rspqpoidlist, OID_AUTO, "dump_count",
3519 CTLFLAG_RW, &qs->rspq.rspq_dump_count,
3520 0, "#rspq entries to dump");
3521 SYSCTL_ADD_PROC(ctx, rspqpoidlist, OID_AUTO, "qdump",
3522 CTLTYPE_STRING | CTLFLAG_RD, &qs->rspq,
3523 0, t3_dump_rspq, "A", "dump of the response queue");
3525 SYSCTL_ADD_UQUAD(ctx, txqpoidlist, OID_AUTO, "dropped",
3526 CTLFLAG_RD, &qs->txq[TXQ_ETH].txq_mr->br_drops,
3527 "#tunneled packets dropped");
3528 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "sendqlen",
3529 CTLFLAG_RD, &qs->txq[TXQ_ETH].sendq.qlen,
3530 0, "#tunneled packets waiting to be sent");
3532 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "queue_pidx",
3533 CTLFLAG_RD, (uint32_t *)(uintptr_t)&qs->txq[TXQ_ETH].txq_mr.br_prod,
3534 0, "#tunneled packets queue producer index");
3535 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "queue_cidx",
3536 CTLFLAG_RD, (uint32_t *)(uintptr_t)&qs->txq[TXQ_ETH].txq_mr.br_cons,
3537 0, "#tunneled packets queue consumer index");
3539 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "processed",
3540 CTLFLAG_RD, &qs->txq[TXQ_ETH].processed,
3541 0, "#tunneled packets processed by the card");
3542 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "cleaned",
3543 CTLFLAG_RD, &txq->cleaned,
3544 0, "#tunneled packets cleaned");
3545 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "in_use",
3546 CTLFLAG_RD, &txq->in_use,
3547 0, "#tunneled packet slots in use");
3548 SYSCTL_ADD_ULONG(ctx, txqpoidlist, OID_AUTO, "frees",
3549 CTLFLAG_RD, &txq->txq_frees,
3550 "#tunneled packets freed");
3551 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "skipped",
3552 CTLFLAG_RD, &txq->txq_skipped,
3553 0, "#tunneled packet descriptors skipped");
3554 SYSCTL_ADD_UQUAD(ctx, txqpoidlist, OID_AUTO, "coalesced",
3555 CTLFLAG_RD, &txq->txq_coalesced,
3556 "#tunneled packets coalesced");
3557 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "enqueued",
3558 CTLFLAG_RD, &txq->txq_enqueued,
3559 0, "#tunneled packets enqueued to hardware");
3560 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "stopped_flags",
3561 CTLFLAG_RD, &qs->txq_stopped,
3562 0, "tx queues stopped");
3563 SYSCTL_ADD_ULONG(ctx, txqpoidlist, OID_AUTO, "phys_addr",
3564 CTLFLAG_RD, &txq->phys_addr,
3565 "physical_address_of the queue");
3566 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "qgen",
3567 CTLFLAG_RW, &qs->txq[TXQ_ETH].gen,
3568 0, "txq generation");
3569 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "hw_cidx",
3570 CTLFLAG_RD, &txq->cidx,
3571 0, "hardware queue cidx");
3572 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "hw_pidx",
3573 CTLFLAG_RD, &txq->pidx,
3574 0, "hardware queue pidx");
3575 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "dump_start",
3576 CTLFLAG_RW, &qs->txq[TXQ_ETH].txq_dump_start,
3577 0, "txq start idx for dump");
3578 SYSCTL_ADD_UINT(ctx, txqpoidlist, OID_AUTO, "dump_count",
3579 CTLFLAG_RW, &qs->txq[TXQ_ETH].txq_dump_count,
3580 0, "txq #entries to dump");
3581 SYSCTL_ADD_PROC(ctx, txqpoidlist, OID_AUTO, "qdump",
3582 CTLTYPE_STRING | CTLFLAG_RD, &qs->txq[TXQ_ETH],
3583 0, t3_dump_txq_eth, "A", "dump of the transmit queue");
3585 SYSCTL_ADD_UINT(ctx, ctrlqpoidlist, OID_AUTO, "dump_start",
3586 CTLFLAG_RW, &qs->txq[TXQ_CTRL].txq_dump_start,
3587 0, "ctrlq start idx for dump");
3588 SYSCTL_ADD_UINT(ctx, ctrlqpoidlist, OID_AUTO, "dump_count",
3589 CTLFLAG_RW, &qs->txq[TXQ_CTRL].txq_dump_count,
3590 0, "ctrl #entries to dump");
3591 SYSCTL_ADD_PROC(ctx, ctrlqpoidlist, OID_AUTO, "qdump",
3592 CTLTYPE_STRING | CTLFLAG_RD, &qs->txq[TXQ_CTRL],
3593 0, t3_dump_txq_ctrl, "A", "dump of the transmit queue");
3595 SYSCTL_ADD_INT(ctx, lropoidlist, OID_AUTO, "lro_queued",
3596 CTLFLAG_RD, &qs->lro.ctrl.lro_queued, 0, NULL);
3597 SYSCTL_ADD_INT(ctx, lropoidlist, OID_AUTO, "lro_flushed",
3598 CTLFLAG_RD, &qs->lro.ctrl.lro_flushed, 0, NULL);
3599 SYSCTL_ADD_INT(ctx, lropoidlist, OID_AUTO, "lro_bad_csum",
3600 CTLFLAG_RD, &qs->lro.ctrl.lro_bad_csum, 0, NULL);
3601 SYSCTL_ADD_INT(ctx, lropoidlist, OID_AUTO, "lro_cnt",
3602 CTLFLAG_RD, &qs->lro.ctrl.lro_cnt, 0, NULL);
3605 /* Now add a node for mac stats. */
3606 poid = SYSCTL_ADD_NODE(ctx, poidlist, OID_AUTO, "mac_stats",
3607 CTLFLAG_RD, NULL, "MAC statistics");
3608 poidlist = SYSCTL_CHILDREN(poid);
3611 * We (ab)use the length argument (arg2) to pass on the offset
3612 * of the data that we are interested in. This is only required
3613 * for the quad counters that are updated from the hardware (we
3614 * make sure that we return the latest value).
3615 * sysctl_handle_macstat first updates *all* the counters from
3616 * the hardware, and then returns the latest value of the
3617 * requested counter. Best would be to update only the
3618 * requested counter from hardware, but t3_mac_update_stats()
3619 * hides all the register details and we don't want to dive into
3622 #define CXGB_SYSCTL_ADD_QUAD(a) SYSCTL_ADD_OID(ctx, poidlist, OID_AUTO, #a, \
3623 (CTLTYPE_U64 | CTLFLAG_RD), pi, offsetof(struct mac_stats, a), \
3624 sysctl_handle_macstat, "QU", 0)
3625 CXGB_SYSCTL_ADD_QUAD(tx_octets);
3626 CXGB_SYSCTL_ADD_QUAD(tx_octets_bad);
3627 CXGB_SYSCTL_ADD_QUAD(tx_frames);
3628 CXGB_SYSCTL_ADD_QUAD(tx_mcast_frames);
3629 CXGB_SYSCTL_ADD_QUAD(tx_bcast_frames);
3630 CXGB_SYSCTL_ADD_QUAD(tx_pause);
3631 CXGB_SYSCTL_ADD_QUAD(tx_deferred);
3632 CXGB_SYSCTL_ADD_QUAD(tx_late_collisions);
3633 CXGB_SYSCTL_ADD_QUAD(tx_total_collisions);
3634 CXGB_SYSCTL_ADD_QUAD(tx_excess_collisions);
3635 CXGB_SYSCTL_ADD_QUAD(tx_underrun);
3636 CXGB_SYSCTL_ADD_QUAD(tx_len_errs);
3637 CXGB_SYSCTL_ADD_QUAD(tx_mac_internal_errs);
3638 CXGB_SYSCTL_ADD_QUAD(tx_excess_deferral);
3639 CXGB_SYSCTL_ADD_QUAD(tx_fcs_errs);
3640 CXGB_SYSCTL_ADD_QUAD(tx_frames_64);
3641 CXGB_SYSCTL_ADD_QUAD(tx_frames_65_127);
3642 CXGB_SYSCTL_ADD_QUAD(tx_frames_128_255);
3643 CXGB_SYSCTL_ADD_QUAD(tx_frames_256_511);
3644 CXGB_SYSCTL_ADD_QUAD(tx_frames_512_1023);
3645 CXGB_SYSCTL_ADD_QUAD(tx_frames_1024_1518);
3646 CXGB_SYSCTL_ADD_QUAD(tx_frames_1519_max);
3647 CXGB_SYSCTL_ADD_QUAD(rx_octets);
3648 CXGB_SYSCTL_ADD_QUAD(rx_octets_bad);
3649 CXGB_SYSCTL_ADD_QUAD(rx_frames);
3650 CXGB_SYSCTL_ADD_QUAD(rx_mcast_frames);
3651 CXGB_SYSCTL_ADD_QUAD(rx_bcast_frames);
3652 CXGB_SYSCTL_ADD_QUAD(rx_pause);
3653 CXGB_SYSCTL_ADD_QUAD(rx_fcs_errs);
3654 CXGB_SYSCTL_ADD_QUAD(rx_align_errs);
3655 CXGB_SYSCTL_ADD_QUAD(rx_symbol_errs);
3656 CXGB_SYSCTL_ADD_QUAD(rx_data_errs);
3657 CXGB_SYSCTL_ADD_QUAD(rx_sequence_errs);
3658 CXGB_SYSCTL_ADD_QUAD(rx_runt);
3659 CXGB_SYSCTL_ADD_QUAD(rx_jabber);
3660 CXGB_SYSCTL_ADD_QUAD(rx_short);
3661 CXGB_SYSCTL_ADD_QUAD(rx_too_long);
3662 CXGB_SYSCTL_ADD_QUAD(rx_mac_internal_errs);
3663 CXGB_SYSCTL_ADD_QUAD(rx_cong_drops);
3664 CXGB_SYSCTL_ADD_QUAD(rx_frames_64);
3665 CXGB_SYSCTL_ADD_QUAD(rx_frames_65_127);
3666 CXGB_SYSCTL_ADD_QUAD(rx_frames_128_255);
3667 CXGB_SYSCTL_ADD_QUAD(rx_frames_256_511);
3668 CXGB_SYSCTL_ADD_QUAD(rx_frames_512_1023);
3669 CXGB_SYSCTL_ADD_QUAD(rx_frames_1024_1518);
3670 CXGB_SYSCTL_ADD_QUAD(rx_frames_1519_max);
3671 #undef CXGB_SYSCTL_ADD_QUAD
3673 #define CXGB_SYSCTL_ADD_ULONG(a) SYSCTL_ADD_ULONG(ctx, poidlist, OID_AUTO, #a, \
3674 CTLFLAG_RD, &mstats->a, 0)
3675 CXGB_SYSCTL_ADD_ULONG(tx_fifo_parity_err);
3676 CXGB_SYSCTL_ADD_ULONG(rx_fifo_parity_err);
3677 CXGB_SYSCTL_ADD_ULONG(tx_fifo_urun);
3678 CXGB_SYSCTL_ADD_ULONG(rx_fifo_ovfl);
3679 CXGB_SYSCTL_ADD_ULONG(serdes_signal_loss);
3680 CXGB_SYSCTL_ADD_ULONG(xaui_pcs_ctc_err);
3681 CXGB_SYSCTL_ADD_ULONG(xaui_pcs_align_change);
3682 CXGB_SYSCTL_ADD_ULONG(num_toggled);
3683 CXGB_SYSCTL_ADD_ULONG(num_resets);
3684 CXGB_SYSCTL_ADD_ULONG(link_faults);
3685 #undef CXGB_SYSCTL_ADD_ULONG
3690 * t3_get_desc - dump an SGE descriptor for debugging purposes
3691 * @qs: the queue set
3692 * @qnum: identifies the specific queue (0..2: Tx, 3:response, 4..5: Rx)
3693 * @idx: the descriptor index in the queue
3694 * @data: where to dump the descriptor contents
3696 * Dumps the contents of a HW descriptor of an SGE queue. Returns the
3697 * size of the descriptor.
3700 t3_get_desc(const struct sge_qset *qs, unsigned int qnum, unsigned int idx,
3701 unsigned char *data)
3707 if (!qs->txq[qnum].desc || idx >= qs->txq[qnum].size)
3709 memcpy(data, &qs->txq[qnum].desc[idx], sizeof(struct tx_desc));
3710 return sizeof(struct tx_desc);
3714 if (!qs->rspq.desc || idx >= qs->rspq.size)
3716 memcpy(data, &qs->rspq.desc[idx], sizeof(struct rsp_desc));
3717 return sizeof(struct rsp_desc);
3721 if (!qs->fl[qnum].desc || idx >= qs->fl[qnum].size)
3723 memcpy(data, &qs->fl[qnum].desc[idx], sizeof(struct rx_desc));
3724 return sizeof(struct rx_desc);